The AM Forum

I finally settled on a linear amplifer for the HPSDR ~100mW 6-160M transceiver.  

Last week I swapped some gear for the PA final unit, the low pass filter assembly and the protection board from a dead FT-1000D.  (May the departed FT-1000D rest in peace)  The previous owner is a vely famous gangsta on this BB.

The solid state PA is good for 200W pep output at about 100mW drive.  The RF low pass filters are relay controlled for each band and will be auto-controlled by the HPSDR with a click of the mouse when bands are changed. Total screen control, no tuning required.

The thermal sensor, mounted on the PA heatsink, will turn on a 12V muffin fan when needed. I expect to use only about 40W pep to drive my tube amp, so it should rarely come on.

The finals run on 30 volts and I have measured my other station FT-1000D to be about -40db 3rd order IMD at 40W. So it is clean enough as a good driver.

I am adding a current limiting circuit for the 30V line that will shut the PA down if too much drive or preset power is exceeded.  In addition, the protection board will generate ALC for high SWR conditions, like any ricebox.

I think borrowing a brick and filter from a DOA ricebox is a good solution for  anyone looking for a low level SDR linear.

Still needs wiring and testing.


73,

T

Pic #2: PA Unit

Pic #3: LP Filters

Those are two very energetic pills. Two hundred real watts PEP? Sweet! Good luck with the project Tom.

That looks like some excellent engineering to "graft" that final stage into linear service with the HPSDR. It'll make a great driver for whatever fire breather you plan to follow it up with.

I made a few component changes to the bandpass board and added the missing band that you wanted right next to the 10 meter filter. Should work FB there now!

You vely, vely funny man, Rob!

Actually those written numbers are the aggregate IQ's of the various bands as estimated by the ARRL.   Your addition will be good for hangover mornings, tnx.

T

Any source for those pcb coax connectors?

Looks pretty damn nice so far Vu. I would maybe think about getting better cooling on that pa board however. Maybe raise the board up a bit on higher standoffs and bend up a little shroud for that fan to get air pressure under that heat sink. My 1000d seems to be happiest at 35 watts carrier but it does get a bit warm evidenced by the little squirrel cage coming on during transmission.

Just a thought.
J

Thanks, Jared -

Yes, I plan to play around with a fan shroud to direct air directly at the heatsink and fins. I realize I should stand the heatsink up so the air flows naturally up. But I will be using it as a 40w pep maximum driver for my tube linear most of the time, so heat won't be a problem. We'll see. I can move it around later if need be.

My station FT-1000D's fan does come on during AM too. It's quite a compact, efficient design, really.

Dave: Which PCB coax connectors do you mean, the tiny 50 ohm ones about 1/8" diameter?  I am short two for the project - I axed HUZ to see if he left them in the rig by mistake. If you mean the 50 ohm phono-plug type, those are easy to make up from audio plugs.

T

Tom,

Those teensie connectors that almost look like they where molded into the end of the coax when it was made are a BEAR to find.....  I've located them twice.

1.  GPS companies use them for interconnects...  Pulled a couple OLD (read BIG) GPS boards apart years ago.  They had them in it.

2.  WiFi Networking.  They are used fairly prolifically for interconnects in routers.

Mouser has them, as well as a few of the SPECIALTY WiFi companies.  They WiFi companies will also have pigtail adapters to make them go to about anything.  I have a couple pigtails here, one for SO connector on the opposite end, one for an N. 

If you want to go the adapter route, the EASIEST (just call and order) would be UniDAPT from RF Parts or your local Bird dealer...  I believe they have an adapter kit, I KNOW RF Parts sells one near their Bird stuff in last years catalog.

First time I saw them was taking apart a Military radio surplus in the 80s.  Was REALLY cool tech then :)



--Shane
KD6VXI

Yes the 1/8" ones.

Thanks Shane.

Glad you clarified -who- estimated the IQ by Band idea... that statement certainly isn't prevalent on 75m SSB!

;D

You will fry that final with the fins down

No problem. I will get some 'L' brackets and mount the heatsink on its side with the fins aligned straight up. The fan will allow a better blast of air this way too.

T

My heat sink is 6 X 12 with 3 inch fins. It gets pretty hot during heavy bias.

Looks like you increased the final cores quite a bit.

Here's the new heatsink mounting. The fan air hits the heatsink fins hard and the air rushes up thru the fins. Vely FB flow.  This is how the sink was positioned in the actual FT-1000D rig.

T

Project update:

The ex-DOA FT-1000D linear transplant is wired up. The 13.8V, 9 volt regulated supply, T/R relays and all DC connections are working.  I still have to install the LP filter interface for the band relay switching. Awaiting parts.
The fan comes on when the heatsink gets warm.

 :o :o :o One word of caution. I almost made a fatal mistake. The Japanese plug for the 13.8V that goes to the PA unit board had the plus wire white and the negative wire red.  I wired it up as I figured it should be but before turning on the power I checked all connections for continuity and proper polarity. Sure enuff, the 13.8V was backwards and would have blown out some series stuff on the PA unit board. OUCH!  Bullet dodged.

So a few more parts need to be added then I'm ready to plug it in for an RF test later this week. I'll dress out the wiring once tested.

The Mercury board receiver, which I pumped 200 watts into blowing the front end last month is back and repaired. The complete HPSDR transceiver is working FB now and ready for the new 200W linear.



T

Project Update:

I put 15V on the 1000D PA unit system including the LP filter. All the T/R DC and RF keying is working.  With 0 dbm drive (1 mW)  I got about 30 watts out.   Looks like it will work FB once I get the 30V on it. Still have related cabling and stuff to iron out first.


One problem I see is the internal antenna relay that feeds the receiver. This RX output shows a slight spike on the scope from the TX on unkey at times. It is as much as 15V.  I will need to sequence this so the TX shuts down completely before the RX comes on.  I once blew out a solid state R-1000 receiver front end because the sequencing was not right. It's easy to get a spark on unkey if we are not careful.

T

The 200W solid state linear amp continues to evolve. See below the small board on top which is for shutdown protection from high swr or excessive final current.  This is the same kit board used in the class E rigs as designed by Steve/QIX. It uses a Hall effect device for current sampling. This is a version of the K4HX/Taylor hybrid.

I also added idiot-proof polarity diodes on the inputs for various low voltage supplies.  Check out the digital volt and current meter.  A small buttom switches from Volts to Amps.

I am still awaiting a control board that will accept the digital band change commands from the HPSDR transceiver. This will switch the low pass filters in the final PA for band changes. All controlled by a mouse.

So far everything is working as it should.  It will eventually run at 30VDC making 200w pep out.  The project's on hold until some more low voltage supplies arrive next week.

I usually keep the wires "haywire" until all testing is finished, then bundle them up neatly later on.

I know what you thinking... "How can dummy like Tom Vu possibly make thing run without blowing up as usual?"  I no know answer to question. Maybe Vu-Konk-Fu vely strong this month.

T

I liked your TMC better  ;)

There's no denying the benefits of an absolutely silent amp, though.  You might need a thermostatically controlled fan.  Both my TransWorld high power solid state amps have them.

John,
That TMC amp is very cool. I was lucky to finally make a deal with a friend who had one. It is unmolested only missing a shield plate. I also bought a pair of new sealed 4CX350s. I was going to buy a Valiant as a medium power rig but the TMC is much nicer. Looked in the stash and found a nice NOS 2000VAC transformer good for at least 1KVA and have another to make the low voltages.
I'm thinking of rebuilding it with new caps and teflon wire.

The TMC will ride again.

Gee Dave I could drive it with an HP8640B and use external modulation input.
That would be very clean. The TMC design is first rate.
If I didn't want the noise I could drop in quad 6146s or a pair of 4D32s on their side.

Hi Johnny,

There is already a thermostatically controlled fan in there. I just borrowed the device sensor output from the amp board and drove the new fan. Works FB with a soldering iron test... ;D  The amp has redundant safety features now, including fuses for RF input, all power supplies, swr shutdown and final PA current electronic overloads protecting both the RF input to the amp -  and final output current.

This amp is almost as big as an FT-1000D box. Those Japanese really were good at squeezing this whole rig in - including the power supply. Amazing how they did it back in 1989.

The TMC certainly is a nicely built amp. However the blower noise is a problem. I tried a pair of 6146's, but the power out was too low in AB1. The amp needs a lot of different voltages, so the power supply takes some time to build. Maybe there will be one at Hoss for you to pick up, Frank.  I saw a complete unit with supply there two years ago for about $400. I tried to buy it but was already sold.

I added extra RF negative feedback to this TMC and it is even cleaner than stock.  I'd say it is as clean as the FT-1000D when the 1000D runs at 40w out.   At full power, the 1000D is no match for the TMC for IMD.

All in all, I'd much rather have the 1000D PA unit for the HPSDR operation due to auto band switching, silent operation and lack of dangerous supply voltages.  (As if there's no dangerous supplies in the shack here)

I know what you thinking - "How can dummy like Tom Vu keep from getting fried all these years?"  I no know answer, but maybe Vu-Kong-Fu vely strong last 40 year.

T

Or "drop in" a pair of 4CX1500Bs as they are only fed voltage. Proportion the B+ up to keep the plate tank as-is. Plenty of dissipation maybe a quiet air system too?

The TMC needs +/-150 volts, High voltage for the final and Zener regulated screen vlotage. Then 120 VAC for the Fan had heater transformer. The only odd thing is the 6CL6 predriver has a grounded screen so you need a negative voltage on the cathode. The tracking predriver and triver tuning is cool.
Wow I have only ever seen one TMC at Deerfield and I now own it after about 10 years.
Dave I looked into dropping in a 4CX1500. One might fit if the socket is sub mounted.
HMMM, I have a socket. I would also add a second common contact on the tank switch. You only have about 4 3/4 inches of height to play with.

Frank, you will need a regulated grid bias too, even though there is no grid current. No sag whatsoever required.  There are relay switching requirements for the voltages too.  Also ant T/R relays needed.  Dave, the TMC you will be getting - I already installed these relays into the amp for voltage and RF ant switching. Just add fixed supply voltages and stir.

Dave, that C1 cap is at the limit for 2KV, so higher voltage is a problem with 1500's.  Maybe a vac variable.  (And a new fil xfmr for the finals)

T

I just checked the 4CX1500 tube height. Min tol is 4.6 inches. With the plate at the bottom of the socket that puts it at 4.7 inches. It might be tuff getting that tube inside the covers. I bet a third 4CX350 would easily fit.

Very interesting. I'll wait to be able to see up close how much fun might be involved.

I'm gonna miss the bipolar sand power drift.....

After all that, you might as well build one from scratch.

That was my plan but looking under the hood it would be hard to duplicate the quality in a 5 1/4 inch rack that runs Valiant power. So I was thinking of rebuilding it with new components and teflon wire. I don't think a 4CX1500 would fit but a third 4CX350 or even some other tubes would fit as long as they are not too tall. I'm thinking an additional loading cap on the lowest band would allow lower impedance operation by just selecting the next higher band on the tank coil. Higher voltage plate tune would be limited to lower C and may not be worth the trouble to install a vacuum cap. I wouldn't push the pi network over 1kw with that band switch.

TMC Swiss cheese...   ;D 


Steve, I was talking with a guy that wants to do an FT-1000 MKV PA unit conversion for his HPSDR. He said it was a 200W unit. I thought only the FT-1000D had 200W PA's and MKV's were 100W.  He sent me this link.  His FT-1000 MKV must be a rarer production unit.

http://forums.qrz.com/showthread.php?252770-FT1000MP-Mark-V-200W

T

Some of the later FT1000 (whatevers) were 200 watt units. I forget which, if I ever knew.

FT-1000MP MKV is 200W, FT-1000MP MKV Field is 100W.  The "Field" runs off 13.8V whereas the straight MKv has a 28V power supply.

Why they made so many different versions of that radio is beyond me.

Oh, so that's why they called it a "field".   It probably belonged out there PW using 13.8V devices... :-)

The 28V MKV  200w PA units might become more common as the years go on and the rigs crap out.  It's really a nice PA strip.

T

Dave,
There is an error in the PA500 schematic. A padder is switched in parallel with the loading cap on the lowest band. Last night I was looking at the switch to see how hard it is to add another common. Looks like I just need to grind a a contact rivet out. I don't know th evalue of the cap but it is a door knob. 

Thanks Frank. We can continue this at Hosstraders over a beer or three.

Sounds good Dave.
I don't like the design of the LV supply. Way too funky for my simple mind.
I plan to use a 300VCT transformer with a bridge to generate plus and minus 150 VDC.

Also, the mark V does 75 watts class A outpoot.

Hay Tom, I have been wondering lately how the HPSDR project is going as well as the FT-1000 rf deck integration. This is my busy travel season so I haven't been on the air much. If it's up and running I'll hear it soon. How about an update and maybe some pictures?

Mike

Gee Dave VW We talked a couple times and totally forgot about the PA500.
I did find a couple 4CX250Bs Friday.

Hi Mike,

I was showing Steve/K4HX the HPSDR and 1000D module linear here this weekend. It's to the point of working as expected with remote mouse control of the LP filters and using a full QIX overload board failsafe system. If the swr climbs too high, drain current exceeds a certain preset value or the heatsink gets too warm there is preventative action.

The only thing left to do is to ramp it up beyond 50 watts to 200w out. I've been testing it carefully to be sure there are no hidden bugs.

We'll post some final shots soon.

TNX for axing, OM.

T

Probs there:  The FETs are expensive, there are two of them, and I think they are around $120 each but worse, they may not be made anymore and are unobtainium now.  The 28 v. supply is some kind of weird switching design and prone to crap out.  Yaesu wanted an insanely high price for replacements so some ops who lost theirs resorted to buying separate Astron 13.8 and 28 v. linear supplies.

OTOH it IS a solid PA.   Once, I ran mine on a DL for testing and did not know the internal fan was not on.  It had gotten hung up in the top speaker's wire.   After 15 or 20 minutes I just happend to put my hand on the heat sink because I make a habit of checking it.  It was hot as hell.  Yikes; I killed the tx and put a fan on the sink and got it cooled down and saved the PA, then got the internal fan working again.  Now when I run a 40 w. carrier with it I have outboard fans on the sink in addition to the stock fan.

The class A option is a bad idea if you ask me--the rig runs way hot, even if you run it at the peak power for class A to get the most efficiency.   It will sit there and tx all day at 20 w. serving as an amp exciter though.

Rob

Rob,

The FT-1000D uses bipolars, not FETS, that cost about $69 each based on a first check:

http://www.datasheetcatalog.org/datasheet/motorola/MRF422.pdf
http://www.pacificsemi.com/search.php?partno=MRF422

Yes, you are correct - I was advised not to run them A1 hot cuz of past failures.  Being bi-polars, they clean up very well when the amp is run at about 35 watts out. That's all I need to drive my tube amp. At 150-200w, they run at the normal -31db 3rd, which is too dirty for me. At 35w they seem about -40db 3rd or so, which is clean enuff for my IPA use.

I use an external 22-48V switching supply. We will see if it causes any switching hash IMD later on. If so, I will go to an analog PS.

I have a big fan on them that comes on with the sensor.

We'll have to see how it works out in the HPSDR service, but shud be FB.

T

I was still  completely brain dead from working overnights and a few other things.
Plenty of time to talk again when I do some work on the thing.