The AM Forum

I have finished up much of the mods to the Ft 102 here. I have noticed that when the AGC is used) on) their is a lot of distortion  ( too much) .  There is something broke in that circuit.

Has anyone run across this problem?

G

I assume you mean on AM. The detector sucks. Replace it with one of the low distortion types found on The AM Window web site. You'll be amazed at the difference. Increasing the time constants on the AGC would probably help too.

All Modes even though I first found it on AM

G,
No probs with AGC here on my 102.

Steve, I found this article, I am going to try the mod.

http://www.securesystemsnw.com/ft/ft-library/FT102/FT-102_Diode_mod.pdf (http://www.securesystemsnw.com/ft/ft-library/FT102/FT-102_Diode_mod.pdf)

Rob, Send me your 102, get off your ass, quit whining and fix the 813 rig. :)

G

HP5082 parts are hard to come by but they were converted to 1N numbers and are pretty cheap. 1N5711 and 1N5712 are the same part. The 5712 has slightly lower junction voltage. This is not a problem at the end of the IF chain. Don't worry about matched quads these things are good stuff and will work great.
I stopped using 1N34s and 1N60s when I got out of crystal radios. They have crappy leakage and very hard to match.
K2CU detector still the best!

OK. Sounds more like an AGC problem then. I never noticed any undue distortion on SSB. Maybe my standard was lower.  ;D

That mod you noted is for SSB only.

Deal, except for the 102 part. ;D
Gary, is there any difference in the AGC action when the AGC button is switched on and off? If I'm not mistaken the AGC is switched by a relay.  I've noticed with my 102 if it sits unused for long periods the AGC behaves strangely until I exercise the switch/relay a few times. And when AGC is switched off there is distortion from strong siginals.

Yes the AGC is selectable. On/Off and fast/slow.

Slow really sucks, about 80 % distortion where fast is about 30%.

I have replaced all the relays in this radio just a few months ago. Actually she works great except for the AGC issue. It seems to me to be a circuit problem.

G

Gary,
That article doesn't show the AGC but I suspect the AGC diode is leaky. I suspect it can't charge the larger slow AGC cap. If you can scan and send me the schematic I can help. BTW the 1N270 is another garbage diode. I'd drop 1N5711s in their place.

Thanks Frank,


Here is a link to the service manual with the schematics. I am guessing that the items in question are on the IF board and the AF board.

http://www.radioamateur.eu/schemi/FT102_serv.pdf (http://www.radioamateur.eu/schemi/FT102_serv.pdf)

G

G,

My Ft-102's compare closely wid the FT-1000D for clean ssb audio.  Sounds like there is a problem in the circuitry there.  Maybe the guys at Circuit City can help. Give em a call.

You're taking the right step with the service manual.

These 102's aren't the easiest thangs to work on, but at least you CAN work on them discrete component-wise. Congrats on getting thru the relay replacement ordeal. If you can do that, then you have the patience to fix the AGC.

If not, you can always send it down to FL for a slight fee... cough-cough.

Let us know what you find.


T

Gary,
There are a crap load of diodes that can cause problems. First watch your S meter in fast and slow AGC. Does it make sense meaning does the meter recover slower in slow mode? Do you have manual RF gain. IF so does the audio sound ok when you back off the RF gain? I need to see if it is the aGC rectifiar or something else. I wonder if you can get at the board with filters on it and tack a wire to monitor voltage. 

Yes Frank , plenty of diodes

I Have full access to the IF board. That is really easy to get to.

Listen to the WAV I have listed below. You can here what the radio is doing.

I used WAV as to not compress the original content. I am using a RE 20 to mic the speaker output of the FT 102 so audio quality was not in the mix.

http://amfone.net/agc test.wav (http://amfone.net/agc test.wav)

G

Gary,
1.Look at the S meter in slow and fast. Do you see a slow release in slow?
2. Do you have an RF gain control. If you reduce the RF gain does the distortion go away?

Yes to number 2

 I see no difference between fast and slow on the meter. I guess that means no.

What I do see is when on slow the meter is showing less signal than in fast. I.E. if the signal is 30  on fast it is 20 on slow. If I reduce the Rf gain it does get better.

Very Good! The problem looks like the AGC rectifier. I'll look at the schematic when I get home and we can go from there. See if you can locate some 1N5711 diodes but 1N914 or 1N4148s or a switching diode should also work.
I wanted to make sure it was really the AGC and not a leaky steering diode.
That would have been a pisser to find. 

G

Have 1N5711s and can stick a few in the mail. Just say the word.

have you cleaned the ---- outta your ears lately????

(Hi Mr. JAY)

G,
I didn't see this in the manual that was posted in one of your previous posts.  This may help in deciphering the schematic.

Gary look at the bottom center of page 51 the IF unit at connector P15-J08.
This is the slow AGC cap. Follow it back to Q12. At the base of Q12 there are a pair of diodes D39, D40 1N270s. I think that is the AGC rectifier. Q12 agrees with Bob's block diagram. Something in that area is a problem. D39,D40 is the first two parts I would check for leakage. Remember anything in that area could be dragging the line down including a leaky cap or bad transistor. Note the RF gain control ties into the collector of Q12. fc

Hi Frank ,

Okay I will start there. I guess the transistor is working, because the Rf gain is working? I could just shotgun the area.

I tried looking for the diodes (5711's) today locally but none in the Syracuse area and those online what me to buy quantities of 500!!

Jay, if you could spare a few I would appreciate it. Let me know what I owe you.

G

How did the balanced modulator mod work out, G?

Did you have any problems getting the resonant transformer peaked, etc?  Audio waveform look clean with 200% DSB peaks?

I've found when set up correctly, 10-15  watts output is about right.

It's certainly a shoehorn trick to fit the HB board in there. I mounted mine on top to the left side.

T

Gary,
I agree if the transistor was leaking the radio would be numb. Bad gain is still possible but I doubt it. Try lifting one leg of each diode and check it with an ohm meter on the diode scale (2Kohm) I suspect one is leaky or open Silicon diodes will work since this level is fairlly high if you want to give it a try.  fc

G

Diodes will ship tomorrow via FCM.

(Hello Buddly - that was quite a QSO the other night on 160!)

The 1N60s near the AGC diodes are the AM detector

G,

If needed I have 1n914s and 1n4148s. I know Jay has offered so consider me a second source.

Jay: Yes indeed. I hope we hear the band in that kinda shape often this winter!
You strapped!!

Thanks Jay on the diodes , what is FCM?

T, I have the AM/FM board so I did the DUQ mods. It works great.

Frank, if I wanted to change out the AM detector what could I use there that may be better?

TNX Buddly

G

Just take the 455 IF out to your downconvertor and use the Flex sw.

Not enough signal Steve, doesn't work off of the jack in the back

But that does pose another question. Where would be a good pick off point for a 455 signal to the Soft rock IF board?

G

A source follower FET stage would allow you to pick it off any place.
I would think around Q11 would be a good spot.
1N60 is a 1N34s little brother. I would think a 914 or 4148 would work fine and have much better leakage. The 5711 will work down to a lower level but you have plenty of signal at Q11. When I did my homebrew RX I found the active detector worked the best and K2CU improved on the one I pulled out of RF design years ago. Softrock is a great spectrum monitor as well as demodulator. gfz

FCM = First Class Mail

Used the K2CU demod on the R-390A before switching to the softrock setup and it worked well. Been meaning to do the AM audio recovery measurements on it to see just how good it is. May get a chance later in the week...been curious for a while now. A Drake 2C is ready for the torture rack tonight.

FB on the source follower. I am going to try that once everything else is set.

TNX

G

The softrock is quite sensitive but low Z input so it must be sucking the signal down. Good FETs to use  2n4416, U308, U309, U310

Ya Know, 1N5711s are like gold to me. Maybe better to try cheap switching diodes to see if you really found the source of the problem. If they work ok leave them. Then save the 5711s for a place that needs them. Just a thought...

Gary,
Measuring those various flavors of Ge diodes on a regular ohm meter (depending on setting) can give you fits.
They will all measure leaky (reverse) by nature (compared to a silicon).  Unless they are physically cracked, they are usually ok.  Betcha its a shorted e'lytic cap sucking down IIRC ?  Q11 or Q12.
My 2 cents

G/L

Jim

I am not Frank but you might consider this circuit that I used in a Heathkit. It greatly improved AM detection.

UPDATE:

The diodes Jay sent me arrived today and I replaced the suspect diodes. It turned out 2 of the four diodes I replaced on the IF board had no readings either way in the diode mode on my Fluke DVM. The other two read fine.

What I find interesting is one diode in each circuit was bad. One in the AM detector and one in the AGC rectifier.

Next are the product detector mods.

Terry, thanks for the AGC circuit. At this point I am going to keep the circuits as stock as possible.

Thanks to all for all the help,the radio sounds much better now.

G

Bud,
When I built my last homebrew RX I went through all of them and settled on the wide dynamic range detector I got out of RF design. K2CU threw it into his simulator and came up with an even better one. I would use Rob's circuit if I was to do it again. Note at 455 KHz the layout and op amp selection more critical than 50 KHz.
Gary, Glad you fixed the AGC. Good thing it wasn't a leaky cap which would have been a pain. Sounds like you need a bag of 1N5711s. I actually figured one diode was open because you didn't have enough AGC action but some.
No action I would have suspected a shorted diode.