The AM Forum

I got this complet POS HW A-7 Amplifer for free.  I would like to use it as a Speech amp.  The Both rigs already have a 8 ohm to Grids Transformer so I need 8 ohms out.  Currently I am using a small Solid state Amp, I really dont care for it and thought this might sound better and be more buzzardly


Here is the idea, This mono amp has two inputs.  I would like to use one for a D104 and the other for my audio rack.  I could then switch back and forth depending on mood. 

Can you please take a few moments to suggest circiut changes to the amp or is it good to go after a recap/cleaning and three wire cord?


Here is a picture of the amp so you get an idea of what it looks like. I will make a screen cover.

http://www.kta-hifi.net/projects/amp_page/a7amp/a7.htm

Here is the schematic.  I thought about a complete redo using one of the speech amps in my handbook..  Oh and I have it working on the variac now.

Good idea.

12A6's in PP are a good design.

73DG

This ones got some nice 12v6s in it.  I am cleaning it up now.  Will replace the caps.  When I have some idea as to what values to use. I have been looking over some speech amp designs.

C

Ok. New orange drops.  Checked the resistors and replaced as needed.  Tubes where good.  Three wire cord also.  She fired right up. Using it for the audio amp on the SP600 now for a test.  She sounds wonderfull.  My only concern is the transformer runs SUPER hot.  Can only hold my hand on it for about 1 second before you have to Draw back.  I tripple checked the voltages and resistance.  Can it be normal to run this hot??

Now to start thinking about conversion to speech amp

C

12A6's are the right tube, a 12V6 gobbles up a lot more heater current.

Also check the cathode bypass cap if not replaced.  Can really raise the plate current if leaky.

73DG

150ma versus 225ma on the heaters.. Big deal?  I can get some 12a6s I guess.  I hooked up a D104 to the input and tested it PA mode.  MAN does it sound warm and rich.

C

Power xfmr should not be running that hot.  Some will run hot after some length of time running, especially xfmrs with a copper magnetic shorting band.

I would recheck everything and I think 12V6s will draw more filament current than 12A6s and probably far more plate current.  I'll check the tube manual....BRB....I'm back.  12A6 draw .15amp fil, 12V6 draw .225amp fil.  Zero signal plate current 12A6 is 30ma and 12V6 is 45ma both with 250V plate.

So the 12V6s are a little heavy for the power xfmr.

Fred

Ok.  They are $2.80 each at Tubesandmore and they are down the street. I will order some in the morning.   

What I really need help with is the inputs. I dont have a manual and cant find one.  I would like to know what the impedance is.   I have found that the one that bypasses the first tube is low.  That much I know.  I would like to convert the first one to run a D104 (probably phono now).

The last thing I need help with is the DAP310 has a Balanced output that right now feeds a balanced input on my Solid state amp.  This thing has no balanced input.  I need help here.  I of course do have a balanced to unbalanced converter but hate to have to use that.



C

From the schematic,  the input impedance of the second tube is 1meg  (look at the grid resistor/volume control).  The first tube (12SL7) is 27K (look at the grid resistor).  You need a much higher grid resistor (5-10meg) for a D-104.  Also, the second stage has some sort of tone circuit, probably for a phono.  You may have to remove some of those parts.

For a balanced input, you'll need a 600ohm to single grid input xfmr.  This is not that easy to come by and usually worth some big bucks if you can find one.  A UTC A-10 is one you can use but I've seen these going for a lot.

My HB xmtr uses a balanced mike input but I have a special UTC mike input xfmr that I pulled from an old RCA studio mixer.  They can be connected for 150ohm or 600ohm balanced input.  I've seen these xfmrs going for a few hundred bucks on ebay.

I think you don't need all the stages of amplification for the output from the DAP310.  I'm not familiar with the DAP but most of this audio stuff runs at line level (usually 1 volt at 600ohms).

I'm thinking that the balanced input xfmr can be connected to the second tube's input.  The first tube can be used for the D-104 and those two stages may be too much amplification but you do have the volume control on the third stage.

Fred

THanks.. I have it up and running on the bench.  Both Inputs now work.  The input 1 (higher gain stage) never worked as someone had a mistake when they built the unit.  I followed each lead down and found the mistake and rewired it.

Input 1 with a D104 sounds tinny and thin with no bass.  I removed the 27K resistor and placed my Eico resistor box there.  10 Meg sounds the best, However, Its still really thin with no bass.  Its not right.  I think its an impedance issue.

Input 2 with the D104 also works and sounds about the same, Thin.  Just has lower gain and requires ALOT of volume knob.

I then tested both inputs with my 600 ohm D104. This one has the k7Dyy Board in the bottom. This gives you 600ohm D104 so you can run the mic into an audio rack or modern radio.

With this mic, The sound is incredible.  Super smooth, clear tone with good low end.  Very hifi.  This mic works with both inputs 1 and 2 and has similiar sound on both. Just  input 1 has a TON of gain.

I do have a 500ohm Merit transoformer, variable ratio, to class B grids.  I also have a small balanced to unbalanced transformer for audio rack gear.

I will have to experiment more.  Its really sounding good with a low impedance input, Just complete crap with a D104.

The Dap310 is designed to drive a BC rig directly via the balanced 600 ohm input.

C

 

You state A7 but you show a A7-E schematic. There was the A7, A7-A, A7-B, A7-C, A-7D, & A7-E. A7, A7-A, A7-B, A7-C used the same output transformer, part #51-13. D and E models used part #51-21. In the D manual, there was an addendum which indicated that 12V6's were being supplied instead of the 12A6's due to supply issues. In concert with that change, the cathode resistor was changed from 470 ohms to 330 ohms. I didn't compare all the various schematics so there may be other changes, although the A7-B schematic only shows 5 tubes.

Pete. This is just an A-7.  Mine has a 470 ohm 2 watt.  Its late, I am tired but I think there is a mistake on the 10 UF cap.  Someone has it from  negative to the chassis and the pos to the cathode.   The schematic shows negative to cathode and pos to the 470K.  In the morning, after coffee, I will sit down and trace this out. 

C

You're wrong, schematic is correct, positive to the cathode and negative to ground.

Fred

More mistakes. The 10 UF was in backwards. I put in a new 10 UF 50 volt the correct way.   

Last night what I was looking at was the 20 UF on the schematic.  That one shows a Negative symbol right on the schematic.  At least thats what it looks like.

C

That's not a minus symbol. The amp uses a 4 section electrolytic can. The symbol shows you which terminal on the can is 20 mfd. The same symbol should also appear on the can adjacent to the terminal. Look at the symbols in the schematic for the three electrolytic caps used in the power supply. Each one of them also has a symbol. They're all part of the same can.

I know about the symbols on the can. I did not realize that minus sign was indicating that.  The amp is playing nice now. I need to come up with a good high impedance input for the d104 now. It sounds great with an audii input. But horrible with tge d104

Mention of the A-7 brings back memories...I used one to dirve a pair of 807s for an outboard modulator I built and used with an Apache back in my WB2ZPS days in high school around 1969-1970.  As I recall, I used a mod transformer from a Navy TDQ VHF transmitter.  The TDQ  modulated the two halves of an 829B in parallel with the 807s.  Don't know if the impedance match was even close to what was needed for the Apache.  I stole the HV for the 807 plates from the Apache and had it wired in to modulate the 6146s  of the Apache in lieu of the built modulator and mod transformer.  Sounded OK, used to play Led Zeppelin II as "test audio" into the dummy load with it, and the mod waveforms lookd pretty good on the old Hickock scope I had at the time, including the Zepplin bass lines from Heartbreaker.

Neat story!.  I was not even alive.  LOL

I found the Bias resistor was measuring at 700 ohm.  I replaced it with a new 470.  Heat is much lower with all the resistor and cap changes. the one output tube is solid blue. Waiting on new tubes. 

Can someone help me understand what the circuit is from the plate of the first half of the 12sl7 to the grid of the second half is for?   I see other schematics online showing just a coupling cap there.  The 0035 cap was very leaky. I replaced it and now the sound is very different.   I removed it and get a BIG increase in gain.  I tried different values for fun.

Thanks

C

If that input was for a phono then that circuit was probably a frequency-compensating network used to balance the way music, etc was recorded on records.  You can remove most all of it,  just leave the grid resistor (680K) and the (.05ufd) coupling cap,  of course leave the 100k plate load resistor.

Fred

How about the 470K in series with the .05?   

TFO came by and gave me a set of metal 12a6s and a super nice little 600ohm balanced input transformer.  I will clip lead that in for a test. This should match the DAP310 nicely.  Then the other input will be for those D104 nights.

C

Remove the 470K, it only reduces the signal much the same way a volume control would.

FB on Dennis stopping by with the tubes and input xfmr.  He was suppose to stop by my QTH with those two BC rigs he has in the pick-up.

Fred

Fred.   I fixed another mistake made the builder and I sorted the 20 UF cap for the 12a6 tubes.  They run hot. Super hot. I guess thats normal. The transformer heat is down now and the sound is clear.  I gutted the entire Phono circiut.  I kept the 470K. There is WAY to much stage gain now.  Even if you flip to that input with no music, Its hot with static. I need to tame it down a bit.

I need to figure out how to match this input 1 to the D104. It sounds like complete crap.  I tested three D104s.  All highs and no mid or low. Sounds Like I am being called to the principals office.. LOL

Keep in mind that this same number 1 input sounds wonderfull with any low impedance mic hooked up.  I tested several. Its very smooth and clear.  It also sounds great with a line out form my ETON SW reciever. 

Currently, On the Grid of the 12ls7, I have a 1 meg to ground and a .05 in series with the mic.   Is this tube just not happy with the high impedance input of the D104?  Should I swap this tube out for another type and redo the front end?

C

For input 1,  D-104 needs to look into the highest resistance possible.  You need at least 5-10meg grid resistor.  Do not use any cap on the input.  Connect the D-104 directly to the grid with the 5-10meg resistor.

I thought there would be too much gain.  You can add another gain control at the grid of the second stage.  It would be in place of the 470K and 680K resistors.  If you don't want to add another control , you can leave in the 470K and reduce the 680K grid resistor to 100-200K.  They form a voltage divider.

Fred

I put the 470K back in. I will lower the 680K when I have time.  For now, The amp is playing great.  Transformer is cool.  Tubes are hot. All voltages are with correct.  The INput 2 is hooked to the output of my Eton XM1 playing FM to a hallicrafters speaker. Input 1 is hooked to a D104 with a 10 meg resistor.  Still no low end. All highs. Not much of anything below 1k. Even with max bass and min treble, the D104 is thin.  I wonder if its a negative feedback issue? 

Tomorrow I will install the fuse holder, OFF/ON lamp and test into the 304Tls.


C

Hey CT,

Just where is that 10 meg resistor?  In series with the mic line, or from grid to ground?

To do it properly, you should've removed the original grid resistor and subbed it.

Just a hint...

73DG

No lows through the 12SL7,  check the two .05 coupling caps, maybe one or both are partially open.  After that, try adding bypass caps to the two 1K cathode resistors.  Whatever you have, 20ufd 25V.

Also, I assuming that you're using a straight D-104, one without anything in the base.

Fred

Zep 1 .... do you really think "Communication Breakdown" was appropriate? *grin*

Maybe you meant Zep II (Heartbreaker was on Zep II)   ;D

You are correct...it was Zep II..corrected the post...

The 10 meg goes from input to ground.  Its not in series.  The input I used was right out of the handbook, the same one I use on my transmiters. c

The trouble was the bypass on the cathode.  Thanks Frank!   I clip leaded a 47 UF/450 cap and each time I hit the Cathode of the SL and SN tubes, the bass came way way up.   All I have is 450 volt caps and I hate to waste them. I will order some 20 UF and 47 UF 50 volt caps today and place them on each cathode of both the SL and SN tubes.  Then, This thing will really sound nice with a D104.

Now to make a simple voltage divider off the 180 volt supply to power some LED on/off indicator lamp.  Then I have to go back to work tomorrow and the fun stops :(

C 

Glad the bypass caps improved the bass LIKE I MENTIONED ABOVE.  Who's Frank??

Too bad you're way out west,  I have some 47ufd 25V caps (NOS) not too old, not sure but, I think about 6000 of them.  I don't want to run my stock too low ;D

DO NOT add bypass caps to the 6SN7.  The first 6SN7 stage has feedback applied to the cathode, that's why there is no bypass cap.  The second stage is the phase splitter.   Just bypass the two cathodes of the 6SL7

FRED

Sorry Fred.  I was thanking you. 

OK.  Will use caps only on the SL7 tube.  I will take some video of the little amp soon.  Need to mount the correct input connectors for XLR input and Two prong D104 to complete the conversion to speech amp.  Also need to make a screen cover.  Then to test it into the 304s!

C 

Here is a picture of the little amp.  Junk Drawer knobs, A neon lamp indicating its on. Dennis donated the 12a6 tubes.  Couple things left to do.  I need to hook up the 600 ohm transformer on input 2 and test it out.  I will clip lead it first. If it works and sounds good, i will mount it.  The D104 input is done. Lots of gain.  Sounds good with some bass boost and treble flat.

Nice looking 12A6's,  I have a bunch of the metal 6V6's and 6L6's.  Don't think I have any 12A6's,  lucky Dennis had them.  Knobs look perfect for the amp.

Can I get some suggestions on negative feedback? I would like to try this as a learning experience. I have read up on this.  The gain of the first stage is very high and I would like to cut it back.  Where and how can I get started?

THanks!

C

1.) The output stage cathode negative feedback mod as shown here will greatly improve its ability to regulate the output voltage over each cycle as your modulator's grids conduct and present a varying load.

It looks very 'busy' due to 4 tubes, but the idea is that the 4 Ohm tap is the CT and 0 and 16 are the ends. Unground "0" and make 4 Ohms the GND and then, 0 and 16 become a balanced winding to buck the cathode currents. The global negative feedback tap can stay where it is but the GFB resistor will probably need adjusted lower (4.7K or a little less) because the GFB voltage will be half with the new connection. The lower, the more GFB, but beyond a certain point it will become unstable.

You might need to use more grid to grid voltage drive from the phase inverter or more gain from its driver, but on this small amp, it should not be an issue because the fixed bias is only -12.5V with Eb=250V. A simple way there is to increase the B+ for the driver and phase inverter to prevent those stages from clipping, if it is an issue.

2.) Also apply fixed bias and get rid of the cathode bias. Bias can be either multiplied up from the 6.3V, or can be taken off one side of the plate voltage winding and rectified.

either or both mods can be applied separately or together.




--> What about the 12A6's plate voltage rating of 250V and this amplifier having 390V on the plate?

12V6's with a 350V rating seem more suited. Is it really running 390V?

I will study the schematic tomorrow.  Thanks for the information Patrick.  I thought about using Negative feedback to reduce the gain of the First stage.

The B+ is 400.  The second supply is around 200 volts.  This is with the factory rectifier. I thought about solid stating the supply, to reduce heat but then the voltages would be way up. 

C

Just looked at the schematic again.  390 volts on the plate seems very high for a tube that's rated to run with 250 volts on the plate.  You can lower the plate voltage by adding a resistor, of some value, between the rectifier and the first filter cap.  Probably 1K-2K figuring about 60ma-80ma total current.  You're using a cap input filter in the PS which always puts a heavier load on the xfmr.  A resistor or choke input filter will drop the B+ and help cool everything down.  Probably better to keep the 5Y3 but, if you go to SS rectifiers then you should add the resistor or choke to the input to the filter.

As for adding FB to lower the gain of the 12SL7,  drop the grid resistor on the second stage to maybe 100K leaving the 470K in series with it.  This will act as a voltage divider and drop the gain of the first two stages.  Adding FB will be more problematic.  You could also replace the grid resistor and the 470K resistor with a volume control, 1/2-1meg audio pot.  You could mount the the pot somewhere on the back or top of the chassis.  Something you can adjust for your D-104 then leave the adjustment as set and use the main volume control to adjust audio level.

Fred

Fred. Thanks.  I will try lowering the voltage.  The transformer runs warm but not hot.  I have some DALE 5w and 10w resistors, I can test some out.

I am going to take your advice and just add a 1meg audio pot. I have a pile of old pots I ripped out of a heatkit O scope.  I will mount one with a knob and label it D104 Gain.  This is very usefull.  I can set the volumn to match the audio rack and this pot to match the D104.

Thanks!

Clark

Clark,

You'll need an audio taper pot.  Not too many of those were ever used in scope circuits.  Scope circuits would have been all linear taper pots.

Fred

Those pots turned out to be .1 meg.  Why would they even mark them like that???  >:(  Why not just say 100k?

I did find a 1.0 meg pot in the junk drawer. Its not marked if its audio.  However, It seems real slow in the first part from 0 to about 200K and then ramps up fast to 1.0 meg.  I will try it in a few minutes.

C

I'll bring that up (.1meg/100K) at the next pot meeting.

That 1meg seems to be an audio pot.

They're usually never marked as audio. Although, some will be marked as linear.

F

Some nomenclature is confusing.  They use both upper and lower case old Greek for lot of electronics notation.

M=million
m=thousand
w=Ohms (sort of a 'wiggly' looking w).

73DG

The wiggly looking w usually refers to 2pi/t   radians/s (IIRC).  The upper case omega is ohms

F

Ah yes, Fred.  In algebraic formulae is does mean that.

Also, for vintage electronics I submit a snippet of an old WECo skizmatic:

73DG

Ah! I see,  no wonder WE went out of business.

Fred

Fred, The 1 meg did not work.  To touchy.  I ended up wth a 20K resistor for the perfect level.  I removed the 1 meg and used a 100K I had.  Its still touchy but works. 

Is this ok??  It strikes me as odd, the resistor had to go from 470K to 20K.  Big change.  However, That input was made for a phono.
 

C

Not sure exactly how you connected the resistors.  The pot should have been in place of BOTH the 470K resistor that is connected to the coupling cap and the grid resistor.

The high end of the pot should be connected to the coupling cap and the low end to ground.  The wiper of the pot should be connected to the grid, nothing else to the grid.

Is this what you did??

Old pots need to be clean with something like WD-40 before you try to use them.

Fred

No. I replaced the 470K with the pot.  Directly. 

From Pin 2 it goes to a .05 coupling cap which goes to Pin 4.  The 470K goes from Pin 4 to ground.  I removed the 470K and placed the pot there. 

The two caps and resistors that where tied to 4 are long gone.


I will redo it after lunch.   

C

Ok. That worked Fred.  Its to touchy still thought. The usefull range of the Pot is only 1 to 1/4.  I need a half value pot so I can get a wider range. Or just leave it.  This stage is super hot with sensitivity

Power supply tests:

At 126volts mains:

450 DC out of rectifiier.
First supply 450dc
Second supply is 420
Third supply is 171

A 2K only got this down to 400

I have a bunch of 10K power resistors.  These got the volts down to 190.  But the rest of the voltages where low and what concerned me was that the volts on the rectifier dropped all the way down to 350.  So I had a 100 volt Drop OUT of the Power supply. This is telling me that the reistor (which runs hot) is pulling down the supply.  Is this expected and how can this be easier on the transformer? The tubes ran cool at 190 volts. The amp seemed fine.  Maybe put two of these 10Ks together to get a 5K and leave it?  Or I can try on the other side of the Filter?

C

I'm not completely following what you're doing with the pot. 

Pin 2 (plate of first stage) has the 100K plate load resistor as always.  Connected to pin 2 (also) is the .05 cap.  The other end of this cap goes to the high end of the pot (nothing else).  The low end of the pot goes to ground.  The wiper of the pot goes to the grid (pin 4) of the 2nd stage, that's it.  No other connections.

If you have the correct type of pot it will work.  You should be able to turn the gain down to zero or all the way up.  You will need an audio taper pot.  You can use whatever value you can dig up (100K-1meg).  If you use a lower value (like 100K pot) you can increase the value of the .05 coupling cap to maybe .1ufd.  This will allow more lower frequencies to be past.  The lower the grid resistor value is, the larger the coupling cap should be.

Fred

I hooked it up the way you told me and it works 100%.  Smooth zero to full blast. Thanks alot Fred! 


C

The resistor in between the rectifier and the first cap is NOT pulling down the supply.  It is doing two things. One, it is dropping some voltage across it, depends on the current and the value of the resistor, Vdrop = RxI.  The second thing it's doing is changing your supply from a cap input filter to a resistor input filter.  This second thing brings down the voltage on the first cap.  To explain,  without a choke or resistor before the 1st cap, the 1st cap tries to charge all the way to the peak ac voltage of the xfmr.  This can be 1.414 times the ac voltage from the xfmr secondary.  When you start your filter with a choke or resistor the DC voltage will be about 0.9 times the ac voltage from the xfmr secondary.

The cap input filter puts a heavier VA (volt amp) load on the xfmr.  The choke input filter puts a lighter VA load on the xfmr.   Choke or resistor input filter allow xfmrs to run cooler.

When you look at ratings for a power xfmr,  it will say how much current the xfmr will handle.  BUT with what type of filter?  Example,  a xfmr might be rated at 100ma with a choke input filter. Now, if you use the xfmr with a cap input filter you will get a higher DC voltage output.  BUT, you must reduce the current drain down from the rated 100ma.  Usually, this current reduction will be about 30%.  So, you should only draw about 70ma from the supply.  This reduction keeps the VA load on the xfmr within its ratings.  Likewise, a xfmr rated at 100ma with a cap input filter can be used with a choke input filter and yield about 30% more current than the rated 100ma.

I'll get back to the resistors on my next post, I've got to yield the computer for awhile.

Fred

I follow 100%.  I just reread the PS chapter in my 1950s handbook. 

For a learning experience, I am just going to redo the entire supply.  There are three voltages needed:

390
360
180


Right now, I have:
425
400
211

Not far off.  I would like to ditch the 5y3. Its just running hot and if I am going to redo this, why keep it? I would like to use a choke. I can rip out the old can cap (not used).  Then mount any choke there on the deck.  Otherwise, I can just use a resistor input, THen adjust the resistors after this one to reach the other voltages.

Parts on hand:

Chokes:
70/350vdc/500ma (golf ball size)
15/1500/1amp (way to big for this but its free.

Caps, 100s,47s,22 at 450, new stock

Lots of 10K, 1500 and 2k  power resistors,

This is lots of fun!

Clark

Clark,

Glad you're having fun,  FB on your parts supply.

Let me know when you get up to more than 1000 xfmrs and chokes, about 250,000 resistors and maybe 100,000 caps.

I'll look at the PS in a bit.

Fred

You may want to keep the 5Y3 for a few reasons.  Your voltage is very near the voltage rating of the filter caps (450V).  When you use SS rectifiers the B+ comes up as soon as turned on.  The tubes are not yet hot enough to draw current.  So, there is very little or no load on the PS.  With little load, the B+ can go way over the 450V and short the caps.  The 5Y3 brings up the B+ in step with the other tubes heating up.  This keeps a load on the supply.

The choke or resistor input filter keeps the caps from charging to the full ac peak voltage, PROVIDED there is a minimum amount of current being drawn from the supply.  That is why bleeder resistors are used in power supplies.

Getting to the resistors,  Use a value that will bring the voltage on the first filter cap down to the 390V or maybe a little lower than 390V  If you go lower like 325-350V, you can adjust the value of the other two series resistors to bring up the two other voltages.  Probably the second voltage can be left at whatever it ends up being, a little lower than the first voltage.  It's the last voltage that you'll probably have to bring up.  I think that resistor was a 100K,  this resistor can be dropped to whatever value will bring up the voltage to the right amount (180V).

After you find the right value series resistor to put between the 5Y3 and the 1st cap, you then must figure what wattage you need.  Two ways,  watts = current through the resistor, squared, times the resistance.  But it may be hard to measure the current.  So, measure the voltage drop across the series resistor.  Wattage = voltage drop, squared,  divided by the resistance.

Once you find how much power is being dissipated in the resistor, you need to use a power resistor 2-4 times that wattage.  You may have to use a number of resistors in parallel to get the right value and wattage.  (2) 10K 10W in parallel will give you 5K 20W.  (3) 10K 10W in parallel will give you 3.33K 30W.  These resistors will get HOT.

Hope this helps,

Fred

Ok. Sounds like a plan. I should have some time later to redo the supply. I need some terminal strips. I am going to hack everything out and start over.  Mount the 5K resistor and then mount the caps, 47/47/22.  Any reason to up this to 100, 100 47?  I have the caps.

The other thing I could do is add a Standby switch.  I noticed yesterday that the B+is 550!!! when you first turn the amp on, then it slowly settles down over about 5 seconds.

I can do this a couple ways, 

1.  Add in a switch (I have about 200 brand new 50s style toggles) that turns the fils on independant of the main, on and off switch.  This would allow me to warm up before turning the volumn knob/on and off switch.  I have LEDs and neon lamps for fil indicator.  Simple and with an indicator lamp would look nice.

2. I could have the new switch operate the B+ stand by/operate and use an indicator for B+

3. The last thought is that I have a really neat looking amperite vac relay(looks like a tube).  I could install this where the old can cap was and wire it up so after the time runs down, it would click on the B+. I think its 5 seconds, Alos with a indicator for B+

C

Are you sure your volt meter is reading correctly??

The 47/47/22 is just right, don't use the larger cap.

No need for a stand-by switch, or any delay relays.

Surprised the 550 volts hasn't shorted the filter cap.

You may need to add a bleeder resistor to keep the peak voltage down.

The best thing to do is put the correct resistor from the 5Y3 to the first filter cap.

Fred

Yes. Even with that resistor installed, The voltage is 470 before the rectifier warms up and volts lower. Its right on the endge. The meter is a fluke.  The voltage is really high before the tube warms up

C

You are measuring the voltage from the negative lead on the filter cap to the first positive on the cap.

Seems a little strange to see that much voltage even before the 5Y3 warms up.  Not sure if that's normal,  nobody ever measures the voltage on the first cap on start up.  Maybe the 5Y3 was still hot from a previous turn on??

If you have the resistors, try adding a bleeder resistor.  Use about 100K 20W.  Connect the bleeder resistor from the SECOND filter cap to ground.  Just clip lead it in for testing.  See if the voltage on start up is somewhat lower.

Fred

I measured right at the Rectifier on the way to the first filter cap.  Voltage is super high at turn on, Then settles down as the tube warms up. 

Here is the plan,

Rip the old crap out.

1. Mount a terminal strip.
2. Run a wire from the rectifier Pin 4 to terminal strip.
3. Frist terminal to 5k 10 watt resistor, then other side of resistor to terminal 2
4. Terminal two is where the first + cap goes.  Other side of cap to ground.
5. Then to a resistor and next cap +, Minus to ground.
6. Then to a resistor and last cap +  Minues to ground.
7. Tap this PS for all three wires leading to the tube stages.

Once that is done, I will adjust the resistors to get  300 volts for the push pull tubes, Then 275 for the second supply and 160 for the last supply.

If the B+ still soars, i will add a Standby switch on the output of the rectifier after the resistor but before the first cap with a Neon indicator lamp.  This is common on Guitar amps to stop cathode stripping and blowing filters out.  Some guitar amps will have 500 to 600 volts before the tube warms up according to google.

Sounds good Fred?   

C

Clark,

Sounds like a good way to be able to adjust the resistor values.  Go ahead with the plan and you can then work forward from that.

Fred

Ok. New PS is working.

411 volts on the rectifier
305 on the first
285 on the second
110 on the last one. 

Need to remove the 100K I have on the last one.  110 is to low.  I would like 190 here.  Amp sounds smooth and there is no hum at all. 

I used 47,47 and 22 caps at 450.    It bleeds down in 5 seconds.   I need to test again to see if the voltage soars up. I forgot to do that.

Neat little amp.  Sounds really nice for a complete piece of trash that was free. 

Really... Thank you for the guidance Fred.

C

   

 

C

Clark,

Those voltages look good.  Just drop the 100K resistor between 2nd and 3rd filters. 

33K got it right on the money.   

Thinking of putting in an VR150 in the open hole in the chassis for the LV supply just for fun. MIght try that later on today. 

C

Clark,

Don't get too creative,  you don't need any VR tubes in this amp.

F

VR tubes look cool though  :o


Its been running for about 45 minutes now in a 75 degree room.    The power trans is at 89 degrees.  The 12a6 tubes are 135f.  The Rectifier is smokin hot at 200f.

The rest of the map is cool.  I will check temps again in an hour just for fun.  The original setup ran the Trans very very hot.  I could not touch the trans without pain.  So as of right now, Its running much cooler.  But it may be heating up still. Its a potted sealed type transformer.

I have to work tomorrow on location so I wont be able to hook the amp up to the rig until tomorrow night at the earliest.  If it sounds good, great, If not, I will hook it to the diode load on an SP600 and let it play 160 meter AM.


C

Is there a possibility that the power transformer has a 105/120V tap and the thing is accidentally running on the 105V tap? What is the filament voltage running during all this super high voltage business?

No the trans has only one primary Pat.  All of the voltages where correct and by design.  Just slightly high due to line voltage.  They spec;d 390 DC.  It had that at 110 volts input. It was a cheap kit amp and did they really ran the crap out of the tubes.


Ok. Its fully warm now and not getting hotter.

Trans was 105F.  Warm.  Remember, Its surrounded by the 12A6s and the 5y3.  So some of that heat is likely radiant heating. 

The 12A6s are running at 135F
The 5y3 is at 162. I think the 200F reading was false.
The preamp tubes are cool.

The 5K power resistor on the output of the rectifier is at 130F. 

I checked the Voltage spike on turn on and its 426DC(at the rectifier tube but before the resistor and caps) the instant you hit the power switch and then fades down to 365 over about 5 seconds.  The big 550 volt spike is now gone.  I would guess this was the resistor input fixing that problem. No need for any stand by switch as my caps are 450 DC rated.

Short of a SS rectifier and a buck boost or lower voltage transformer, I dont think we can get better then this!

Fred your idea worked perfectly!

C

Clark

I would leave in the 5Y3.  You'll only raise the voltages again.  If your voltages were too low you might do the SS mod.

I built (HB) an amp very similar to your amp with 6V6s back in 1964.  Worked great and still works.  My voltages were a little on the low side so I SS the rectifiers many years ago.  Still working but I may put the 5Y3 back in.

OK on the temp measurements.  I never did any myself except for touch measurements.

Small xfmrs will run warm but shouldn't be hot.  Tubes run hot, especially any of the power tubes and rectifiers.

Fred

I agree.  We took 75F off the Transformer.  So it has to be happier.  I will leave the amp as is. Time for on air testing tonight.

C