The AM Forum
March 28, 2024, 03:41:56 PM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: 1 ... 3 [4] 5   Go Down
  Print  
Author Topic: ARC-5 works great on AM...  (Read 93419 times)
0 Members and 1 Guest are viewing this topic.
WB3JOK
Member

Offline Offline

Posts: 637



« Reply #75 on: February 09, 2015, 09:21:19 PM »

I went whole hog. A pair of 811A GG is what I use if I need to "Reach" with my ARC5. 12v heater supply fed down the middle. My own concoction.
don

How did you match the very low-Z output of the ARC-5 to the input of the 811A's?
What plate voltage/current is your supply, and how much power are you getting out of that configuration?
Logged
VE3LYX
Member

Offline Offline

Posts: 769


Crystals are from the stone age


WWW
« Reply #76 on: February 10, 2015, 09:26:44 AM »

I made a simple antenna tuner for matching other stuff in the shack and used to to make the arc5 happy feeding the twin 811As which have an untuned input BTW. Power out appears to be 100watts or so. I don't trust in shack meters because stray RF can throw you. I use the old school dummy loads you can read by. No power no light , period. (And no I have never not once ever had a QSO on one of them.) I tune for max and then switch to antenna (longwire) and only trim if absolutely required. Usually it isnt. I forget the HV. Probably 1100 to 12oo volt keydown. I use indicator bulbs rather then meters cause I can barely read the meters anyway. This station is old school so perhaps may not interest most but it works and that to me is very interesting.  Normally I use the 811A deal with my DX60B without the tuner.
don
The twin 811A build was on here somewhere and was a struggle at first but is a very robust trustworthy addition to my shack. It was my own design based on others I had seen over the years. China tubes and I have pushed them into the orange several times but not with the ARC5. Many here who have worked me have worked me on this combo either the arc5 or more likely the Dx60B on this linear. I use it in marginal conditions . I really do not care for big power. Why I don't know. I just dont. Of course I ran a six cyl dragster for many years too. Must be an underdog thing.
Logged

Don VE3LYX<br />Eng, DE & petite Francais
WB3JOK
Member

Offline Offline

Posts: 637



« Reply #77 on: February 11, 2015, 09:48:17 PM »

Don, sometimes it's just fun to do more with less  Cool

Got back from my trip and tinkered with the PW'er some more. The modulation looks good, but when listening to my carrier (into the dummy load) on the receiver with the antenna disconnected, there is an unacceptable level of 120 Hz hum with the audio gain at zero. It's not very noticeable on the scope but it sure is audible... The problem is that the 6AQ5 screen modulator tube is fed from the plate supply so any ripple on that supply also directly modulates the PA screens  Tongue
 
So I either need to improve the power supply (which I had been thinking of redesigning anyway, as posted earlier), or put a good filter between the B+ and the plate of the 6AQ5. I don't have a small low-current 5 or 10H choke, or any room to mount one. Perhaps an RC filter since the current draw is only a few milliamps. Either way I'd need another electrolytic capable of withstanding the full B+(near 600V currently). Hummm...  Huh
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #78 on: February 11, 2015, 11:08:02 PM »

I found I need a good stiff voltage feed to the screen modulator.
Plenty of capacitance.
Logged
WB3JOK
Member

Offline Offline

Posts: 637



« Reply #79 on: February 12, 2015, 03:47:28 PM »

The existing plate supply would only drop from 600V idling to 580V but there was something like 20V p-p ripple on it at full carrier.

I found a decent Stancor power transformer in the junkbox labeled 270-0-270 @ 120 ma DC. It's more like 600 VAC unloaded with my relatively high line voltage, since it was designed in the days of 117 VAC line. I used the 5 VAC rectifier filament winding to buck the line, which helped some. That should give me plenty of plate voltage with the existing bridge rectifier - maybe too much!

Anyway today I removed the unnecessarily large 250VA control transformer and installed the Stancor on my compact chassis with barely enough room for it and a small 50VA control transformer also wired backwards in FW CT with a couple of 1N4007's to supply the oscillator regulated B+.  The Stancor's 6.3V filament winding is now feeding the 6AQ5, which allowed me to remove another transformer from under the chassis.

No more "economy" power supply circuit. Now I can, if need be, increase the filter caps on the plate/screen supply without affecting anything else. Stopping to eat, then smoke test... with a Variac  Wink
Logged
WB3JOK
Member

Offline Offline

Posts: 637



« Reply #80 on: February 12, 2015, 05:14:00 PM »

Passed the smoke test (except I'd forgotten to hook up the primary of one transformer)  Roll Eyes

Now I have about 700V on the plates, and considerably less hum, since the HV filter caps (two 100 uf in series, or 50 uf) are not dividing across the LV side caps of the former "economy" supply any more. My effective capacitance was only 26 uf, and now of course it's 50 uf. Not surprisingly it reduced the ripple! I may tinker with an RC filter to the 6AQ5 since there's more than enough headroom for voltage drop in the R.

Tuned for largest cleanest peaks into 50 ohms, there is now about 9 watts carrier, and 64W PEP on voice peaks  Cool
That's enough work for a pair of 1625's I think. They aren't very hot, either.

Nobody on 40m this afternoon to try it out on  Sad Sometimes there's AM activity on 7160. There's a strange broadcast station around 7290 - in English but with an indefinable accent I can't identify. Chinese broadcasters?
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #81 on: February 12, 2015, 07:18:40 PM »

Too much audio, unload the final some to get more carrier.
If you get 64 watts pep you can run 15 watts of carrier.

Logged
WB3JOK
Member

Offline Offline

Posts: 637



« Reply #82 on: February 12, 2015, 07:46:29 PM »

Sure, I can load it (and adjust the screen mod bias) to run more carrier, but do I want more carrier? The modulation is what carries the information, in the sidebands... and at the current settings I get the biggest peaks on the scope.

Won't I have a more audible, intelligible signal with the highest upward modulation possible without clipping at baseline?
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #83 on: February 12, 2015, 10:05:03 PM »

4 times carrier is about 100% modulation, and if you go past that you can sound distorted in most receivers.
Unless its an sdr with a sync detector, most receivers distort badly over about 150% positive modulation.
Many normal transmitters will not do 4 times carrier without a lot of work, plenty of audio power, separate power supply for the modulator, tweaks to the screen modulation in tetrodes, etc.

I can adjust my screen modulated rigs to run silly low carrier power and super high peak power, say 100 or 200 watts of carrier and 1500 watts pep, but it does not sound good, and can start being dsb with reduced carrier which is what it actually is...
Logged
WB3JOK
Member

Offline Offline

Posts: 637



« Reply #84 on: February 12, 2015, 10:12:03 PM »

When you say "four times carrier" you do mean four times the power, right?

Edit: obviously... 100% mod goes from 0 to twice carrier amplitude which is four times the power  Roll Eyes

I will try adjusting the audio gain, loading and screen bias for twice the carrier amplitude on peaks... until I get this thing on the air, it's pretty hard to tell what it's going to sound like (ever tried to listen to yourself while you talk?)   Wink
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #85 on: February 13, 2015, 08:06:17 AM »

I listen to myself in the mod monitor all the time, I want to know if something happens to the audio.
Once you start doing that, its like an SDR bandscope, its hard to stop using it.

AM modulation: 100% modulation is 0 and 2X the plate voltage on the final tube.
When you double the plate voltage, you get 4x the output power.
So if you ran 15 watts of carrier power, you should have 60 watts PEP.

In many plate modulated transmitters, the modulator power supply sags a bit on peaks and can not deliver full peak power, or the final tubes run out of emission to handle the peak current you would have with 2x the plate voltage, sometimes the screen gets in the way in tetrodes, sometimes its a grid drive/bias problem, but many rigs may only do 3x the power output.
My stock 32V's would only do 3 times the carrier power, mostly because of the screen voltage setup on the final tube.

Anyway, its great you are able to get such nice peaks, but more is not always better.
The loading control sets the carrier to audio level, you can adjust it to give 60 watts carrier out with only downward modulation or 5 watts carrier with 60 watts pep output.
If the values in the output were right, you might be able to adjust things to get one watt carrier and 60 watts pep, it comes down to the tank coil and loading cap..
For the lowest distortion, you would want about 4 times the carrier power as a pep reading.
Measuring average power, it should stay steady, not go up or down.
That is another way to tell things are adjusted right, you get 4X the carrier power and the average power does not change. Not enough loading and the average power goes down, too much and it goes up.

So, if you start from the point of maximum power output, more loading (loading cap more open) gives less carrier power output and more peak power.

Logged
W2VW
Contributing
Member
*
Offline Offline

Posts: 3489


WWW
« Reply #86 on: February 13, 2015, 10:15:17 AM »

4 times carrier power only exists for sinewave program. If there is sufficient bandwidth that rule of thumb can be exceeded.
Logged
WB3JOK
Member

Offline Offline

Posts: 637



« Reply #87 on: February 13, 2015, 01:11:17 PM »

I set it for more carrier (maybe 12W) and still 60W PEP on voice peaks, and was able to get into the Noon Forum on 7295 with my "All ARC-5" (both the rx and tx) and a homemade TR switch Grin

Jon AC0OQ in Iowa, today's NCS, reported I was coming in S9, loud and clear, but with slight audio distortion, although it looked OK on the scope. My mic is an old Shure hand-held so definitely not hi-fi... He was also using a notch filter for heterodynes and a couple other guys got the same audio report, so maybe it's on both ends.

Right now the "leenyar" has the filaments (only) lit, cooking for half an hour or so (I've read about power tubes flashing over if they've been dormant for a long time, and it's been close to ten years since I fired it up).



* P02-13-15_11.02.jpg (201.35 KB, 1280x960 - viewed 862 times.)

* P02-13-15_11.11.jpg (147.01 KB, 1280x960 - viewed 863 times.)

* P02-13-15_11.11[2].jpg (192.88 KB, 1280x960 - viewed 839 times.)
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #88 on: February 13, 2015, 01:29:01 PM »

That amp looks nice, just plug in some 4-400's.

If you get reports of being distorted, you need to find a way to listen to the audio and adjust things for the best sound.

Once you get the screen voltage and audio levels in the ballpark, you can adjust the grid drive/bias and the loading control.
Did you add the feedback from the screen modulatior output to an early stage?
That helps a LOT. Makes things less dependant on grid drive/bias.
 
Logged
WB3JOK
Member

Offline Offline

Posts: 637



« Reply #89 on: February 13, 2015, 03:13:02 PM »

Don't have enough iron (plate or filament)... my Thordarson plate transformer is rated for 900W CCS and there are two 5V, 14 amp filament transformers. I designed this amp with the ARC-5 in mind, and to use 4-125A's because they are cheap and readily available (I think four of them cost less than one 4-400). At least they were ten years ago!

I let the filaments and fan run for an hour, then switched on the HV. 3.4 KV with a light bleeder, 3.2 KV with 100 ma cathode current (all four tubes) with 400V on the screens, some orange on the plates at 80 watts per tube which is normal. No magic smoke release.
 
Couldn't remember how I wired the 24V T-R relay and tired of playing with radio today - not good to be tired or impatient when working near 3 KV supplies with 56 uf filter caps. Shocked Maybe over the weekend I'll get back to it.
Logged
WB3JOK
Member

Offline Offline

Posts: 637



« Reply #90 on: February 13, 2015, 08:07:36 PM »

Although, after looking at the specs, I could plug 4-400A's into two of the sockets and leave the other two empty. Still don't have enough plate transformer to support full legal output.

Anyhow, I sorted out the TR switch mess and got everything to key up with the mic button on the ARC-5. Unfortunately the amp is not working (well, it's putting 0.75 RF amps into the 50 ohm dummy load, and the ARC-5 by itself makes 0.5 amp)  Tongue Meters say the voltages are normal (no screen current on the panel meter, but that's not unusual in AB1 with only 450V on the screens).

Cathode current in "Tune" with no input is about 90 ma with everything warm, and 120 ma with the ARC-5 driving it. Plate tuning is very peaky, and most strangely, with the loading padding caps switched out, the output power is still rising as I reach max capacitance on the loading vacuum variable. But the output drops with the first padding cap in, no matter what I do with the variable, and even worse with more caps!

There should be. as usual, a broader peak of plate tuning with increasing loading, and of course more output. So something is definitely screwed up. Maybe a bad solder joint somewhere. Got to scope it and see if the drive through the TR relay, swamping resistor, and 4:1 unun is actually making it to the grids, to begin with. Working anywhere near that HV supply still scares me though...

Edit: After studying the 4-125A data sheets again, there should be a few ma of screen current with the input swinging the grid about 100V p-p. Screen current could be zero or even negative with no signal or severe underloading. Also 120 ma cathode current isn't near enough, should be more like 200 ma at full output. I'm betting there's an open in the tank circuit. We shall see...
Logged
VE3LYX
Member

Offline Offline

Posts: 769


Crystals are from the stone age


WWW
« Reply #91 on: February 14, 2015, 09:19:31 AM »

I remember fighting through some similar deals with my twin 811A amp. I could orange the plates but got next to zip out. I had to revamp my plate choke and even remake my parasitics. It took very little difference to cure. I also experimented with link output vrs PI and settled on PI . Getting enough C in the load cap was a big deal. 1000pf just didn't cut it. But you are a doer and I know and can see you will sort it out. Till then,
bTW picked up a pair of MkII no 19 wireless sets. One is supposedly NOS unused. No cables but I didnt have any for the Arc5s either. Output is less then an ARC5 as only one 807 as opposed to two (1625s) however it should be able to excite the amp enough to enjoy it. They are coming on the bus from Northern Ontario sometime today. Meanwhile to your success!
don   
Logged

Don VE3LYX<br />Eng, DE & petite Francais
WB3JOK
Member

Offline Offline

Posts: 637



« Reply #92 on: February 14, 2015, 01:09:19 PM »

yeah, some of it was operator error (have three taps on my toroid tank coil and only the middle one gives decent output). Also it seems as though the vacuum cap on the loading side of the pi-net works "backwards" (clockwise rotation of the shaft decreases the capacitance). Since it has a ceramic body I can't see its plates.

Also I should have known better, that adjusting the loading cap for optimum output with the unmodulated carrier is NOT the correct loading at full output  Lips sealed

Furthermore, I may have been planning to use this amp with the Heathkit single-banders I used to own - there is a T-pad that I tucked under the chassis and forgot about, before the bifilar step-up unun (4:1 Z, 2:1 voltage) with a 200 ohm swamping resistor.

I jumpered around the pad and eureka! 280V p-p carrier into 50 ohms (which is 200 watts).  Maximum on voice peaks is just about double (540V p-p) which is 730 watts PEP  Cool Not sure yet if the ARC-5 or the amp itself is the limiting factor.

AM Noon Forum on 7290 was just ending, naturally (and being interfered with by on- or nearby frequency SSB activity as usual). There was a good round-table going on 7160, mostly 5-land, but I only tried a couple of times to break in barefoot and they didn't hear me. I'll see if I can try the noon forum again in the next day or two.

Don, what's a MKII? Pics?

Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #93 on: February 14, 2015, 03:58:47 PM »

I was on 40 today till about 12:30, 7295.
If you hear me on, call me.
It would be fun to work screen mod to screen mod rigs.

What did the 4-125 plates look like at 200 watts out?
I would think you would want to run them at about 150 watts out and 600 watts pep..

Logged
WB3JOK
Member

Offline Offline

Posts: 637



« Reply #94 on: February 14, 2015, 05:18:23 PM »

Didn't hear you but it was just about that time or later when I got on the air... will listen to 7295 and 7160 tomorrow or Mon. morning if I get a chance.

Plates were moderately orange which I believe is normal behavior for these tetrodes. Even at no-signal idle, 85 ma at 3250 V = 280W, or 70W per tube, there is still a dull orange glow on the plates.

I have a good sized muffin fan blowing across the bases and envelopes whenever the filaments are on, and large finned plate caps, which should keep the seals out of trouble. The RCA and Eimac data sheets for 4-125A/4D21 claim that convection cooling alone (sockets mounted in the clear to allow circulation) is sufficient at the full 125W dissipation if below 30 MHz, and that no more than 2 cfm of air is plenty at higher frequency max ratings. With four tubes mounted close together I wanted a bit more cooling than that - these tubes are cheap but not that cheap.

I have not seen the meter show more than 200 ma at 3200V = 640W DC input, and that's with 200 W coming out, leaving the remainder safely under the 500W total max dissipation  Cool

It'd probably be a good idea to turn the wick down just a bit, though, in the interest of extended tube life. 1 or 2 dB shouldn't make much difference. But going from 10W to 150W (12 dB) sure will!
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #95 on: February 14, 2015, 08:47:37 PM »

Normal operation is a nice bright glow, no problem doing that.
What you want to do is (somewhat) match the carrier power to the peak power.
Things can be adjusted to give about 4 times carrier power on voice peaks.
200 watts carrier would be 800 watts pep.
If you can adjust things to do that with the tubes below white hot you are golden!

If things do a maximum of 600 watts pep you should run the carrier at about 150 watts.
You do not want to flat top, or run 100 watts carrier and 800 watts pep, the least distortion will be around 4 times carrier power.

Its going to be a mess adjusting things, the screen modulator (output voltage, audio level), the transmitter grid drive, loading, the amplifier.


Logged
WB3JOK
Member

Offline Offline

Posts: 637



« Reply #96 on: February 14, 2015, 09:24:58 PM »

Heard some guys on 7295 this evening and joined in the roundtable, then continued a good QSO with Tim KC9QJE. He told me my signal sounds good once I turned the audio gain down a bit, which is good news. Still S9 barefoot, but nearly S9+20 with the amp so it'll be really helpful in marginal conditions... that's what it's for, after all.  Smiley

However, he reported that on his Softrock SDR display, my signal is about 3 KHz wide, with or without the amp, but that nearly all the energy is in the upper sideband, hardly any lower sideband.

I'm trying to figure out how that's even possible. When screen modulating a MOPA transmitter, I would expect both sidebands to be equal and 1/2 the amplitude of the carrier at 100% mod. Could there be a tuned circuit in the ARC-5 that is tracking slightly high - but it would have to have an impossibly high Q to reject the LSB less than 6 KHz away at 7.295 MHz. What's going on??  Huh
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #97 on: February 14, 2015, 10:55:29 PM »

I have seen that on some transmitters.
I have been told its phase modulation, audio getting into the vfo.

3 KHz is very narrow, what would limit the audio?
What do you have for screen bypass caps on the 1625's?
Should be about 500 pf I would think.
 
Logged
WB3JOK
Member

Offline Offline

Posts: 637



« Reply #98 on: February 14, 2015, 11:32:56 PM »

3 KHz is not that narrow if it's only one sideband though - would be 6 KHz with both sidebands, right?

I'm using an old handheld Shure "CB-10E" mic. Very heavy cast aluminum case and heavy coil cord. I'll hook it up to my HP 3580A audio spectrum analyzer and see where the high-freq rolloff is with my speech patterns. I suspect it's 2 to 3 KHz by design. SSSSSibilance. Sibilance.  Grin

There is a small LC filter I added to the input, to keep RF out of the audio. It's 30 uh series and .0022 to ground. Now that might be rolling off the incoming audio - don't know the Z of the mic. Really I'm not looking for hi-fi audio anyway - why run highs up to 5 KHz and use even more bandwidth?

No other low-pass elements between there and the PA (except the screen bypass caps). Those caps are original to the ARC-5 transmitter (.002 uf with a series 620 ohm resistor).  A 0.002 uf cap at 5 KHz has a reactance of 16k ohms. Shouldn't be doing too much shunting with the cathode follower driving the screens...

Phase modulation from audio into the VFO? OK, but how's it getting in there... The VFO is now regulated by two OB2's at 216 volts nominal, and the only other load on the 280-300V unreg supply that feeds the VR tubes is the first stage of the speech amp, which pulls less than 1 ma and is RC decoupled anyway.

Or is the oscillator pulling with the modulation on the PA? Several KHz though? I'm still not clear on how that would eliminate the lower sideband anyway...



Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #99 on: February 15, 2015, 12:14:41 AM »

3 KHz is ok if its only one sideband.
I do not know how it does it, but if the vfo gets pulled with modulation, its in one direction.
Most times I have seen it its the lower sideband that is thin or missing.
Logged
Pages: 1 ... 3 [4] 5   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.064 seconds with 19 queries.