ft-901dm ham radio repair.

[retrolefty]

"IF transmitter is OK without the ALC loop closed. It points a big fat finger to the loop it self. Carefully look at the circuit around driver and how the RF voltage is sampled. It may be you have a resonance at 7 MHz creating far too much ALC applied. "

Yes Mr Simpleton. The transmitter does appear to be fully working without the ALC loop connected (or with PA valves removed). We did discuss resonance around the ALC rectifier circuit some posts back. But with no result. Components were tested/changed.

I still believe this is a resonance related issue effecting the ALC detector. It could even be a simple Valve type issue. After reading some docs about various types of 6146. It seems the 6146B is not always compatible with earlier 6146 / A valves. It also states some of the problems in changing to 6146B's are 'differences in inter-electrode capacitance and spurious parasitic oscillations..' hmm, sound familiar?

 Sounds familiar. I have owned a Kenwood TS-520S hybrid forever it seems (idle in a closet) that used the Japanese version of the 6146/6146A, S-2002 I think I recall. I too recall reading that mixing between 6146/A/B some of the older hybrid transceivers can be tricky.

[voltz]

"Voltz, it's my understanding you think the waveforms as presented on tds540 to be clean, is that correct?"

Yes. But let me define 'clean':
The jiggle you are seeing is the expected result from the mixer. Its not clean in the classic sense of the word, ie a pure sine wave, but i believe it is correct and should not cause a problem to the ALC.

If i'm wrong but your radio gets fixed, its still a great result because your radio got fixed!

[drforbin]

I agree....Just wanted to clarify

[voltz]

Your plan of attack sounds good. Divide and conquer.

But, what types of valves do you have there at your disposal? Final PA valves. Can you change them. And compare ALC levels.
The biggest thing that stuck in my mind was your finding about pulling out the finals and curing the ALC. Thats just weird and we're back to finals again. See my 6146 types post earlier.. What do you think?.

[drforbin]

Well, there is a difference between the a/b/w
The B tubes should be used. The W version is a militarized version(Us Americans love our military you know).
I have both the B and W versions available. Both produce the same results. 

[voltz]

I read that B and W types are basically the same with different branding letters. But the earlier 6146 and 6146A are more stable and not the same. Are we sure this radio was designed for B types?

Im getting my information from here:
http://www.foxtango.kc9foz.com/FT101ZD-901-902/FT-901DMsg.pdf

page 25 of the pdf.

EDIT: no, B and W are not the same, after reading again, but W types can be any type. Only the date stamp gives it away.

[drforbin]

This is from service manual.

[voltz]

Indeed.
But something is wrong here. Maybe a version change at some point? Its sounds 'out there' doesn't it.

[drforbin]

I know...that's why I needed (need) and appreciate you guys help.
It's a nutty..will post SA analysis a bit later.

[voltz]

Seems like there were at least two versions of the 901dm. An early version and a later one. Yours appears to be an earlier version (looking at your trimmer board colour). But i cannot find any references to what finals were used in the earlier type. Can anyone comment on that?.

[drforbin]

yes. you are correct, mine is an early model.  The finals and driver are the same.
You can download the ft-901 survival guide from tango foxtrot and it will  detail the difference between versions of all components. 

[AF6LJ]

Here is link for manual.

http://www.foxtango.org/ft-library/FT-Library/FT101ZD-901-902/FT-101ZD-901-902.htm

As for peaking fine. But I have done that alrdy multiple times. Peaking I mean.
I did not use the SA. In fact it's not possible to peak the 40m with the ALC connected because it's fighting you the whole time. I do agree with you. I believe there is nothing wrong with the alc circuit.
It's very simply and I have checked all the possible fault modes.

You said the signal looked dirty, I request again, which part?

The vco input to rf unit, the signal input to rf unit or the resultant output of rf unit?

The PA output looked real clean to me but it's been filtered alot by that time.


thanxs for your time.

The signal going into the driver should have been cleaner than it was.

[AF6LJ]

From reply #43:

"ok..just used SA on PA. I used a 30Mhz span to capture at least two harmonics.
I'm getting a perfect peak at 7Mhz nothing else."

You are looking too far out.
You want to look within 5MHZ of the 7MHZ signal.

guys, its clean... and tuned. Otherwise there's no way the TX would achieve 100 watts at that frequency. (with ALC disconnected). no?

Whatever is coming out of the RF mixer (jiggle on scope) is just the result of mixing. You can see 'jiggle' on your scope at 20m too. And is then filtered out at the following driver stage. Its all working as it should! Power and frequency wise. Its just the ALC is 'seeing' something wrong. Thats the mystery. That is my current thinking of it. It might all change when we see this SA result!

Not if you are looking so far out.
See above...\

[AF6LJ]

Well, there is a difference between the a/b/w
The B tubes should be used. The W version is a militarized version(Us Americans love our military you know).
I have both the B and W versions available. Both produce the same results.

If you are getting a hundred watts out on the other bands your tubs are good, you are barking up the wrong tree.

[AF6LJ]

Indeed.
But something is wrong here. Maybe a version change at some point? Its sounds 'out there' doesn't it.

Something is wrong but it is not in the PA.

[drforbin]

Ok guys...
here is SA! What do you all think?
span is 5 Mhz to 50Mhz
resolution BW 1 Khz

[voltz]

Yep, pretty much what you expect to see from a wide band mixer output with no filtering. And it would look similar on all bands.

Looking at the RF unit circuit PB1702, there is no filtering at all to terminal T2 out (pin 6). Its totally wideband. Only when it hits the driver board does filtering start. Which is what i was saying some posts ago.

Ok, since the valve type theory and various others have so far lead nowhere, i think AF6LJ's suggestion of getting in closer to 7Mhz is a good one.
Say take a look within a few Mhz only +- of 7. If there is any unwanted energy next to it, you would need to get in close to see it.

[EDIT]
But looking at your SA, you can actually see the IF at 9Mhz and its very low in comparison to the wanted 7. So thats correct.

[voltz]

It would be interesting to see SA at the RF Out phono jack a the back of the unit. This is a convenient connection to the actual grids the ALC is monitoring. If there is a tuning issue or resonance happening, surely we would see it there. It should be quite pure 7Mhz at that point. After driver filtering.

[AF6LJ]

A few things that need to be said..
The mixer output is not designed to work into 50 ohms.

I had asked to see close in to the carrier (+ / - 5 MHZ)
Well since I am not getting the data I need to see.


Let us all know what you find when you finally fix it.

Take Care
I am done with this thread.

[drforbin]

AF6Lj,

I'm sorry. I really donot remember you stating you wanted a up close of carrier.
Thank you for trying to help though.

[AF6LJ]

AF6Lj,

I'm sorry. I really donot remember you stating you wanted a up close of carrier.
Thank you for trying to help though.

First you need to understand that none of those cables in that radio represent a 50 ohm system. So when you plug one into a a 50 ohm device such as a spectrum analyzer the readings are meaningless. Plus you risk putting DC on the input of the SA, that's bad.

With that said;
Build yourself a sampling loop.
Take a BNC cable and cut the connector off one end and strip back the outer jacket about 2",  and all but a quarter inch of the shield, slide a piece of heat shrink over the cable.
Strip a quarter inch of insulation off the center conductor and solder it to the shield forming a loop. Bring the heat shrink tubing up over the joint to insulate it and shrink the tubing... Now you have a sniffing loop.
This is what you use to sense RF in circuits where your gear will load the circuits down

What you are looking for are relative signal levels, this loop is good enough and you won't load down the circuits in question and you won't damage your SA.

The only signals you are concerned with are those within a few MHZ of 7.0MHZ, so looking at the second, third, harmonics will do nothing to lead to finding the fault.

Here is what you want to do;
Tune the radio up on 7.0 MHZ with the ALC connected. (that's right)
With your sniffing placed near the bottom of the driver tube look at the spectrum display. Set your frequency span to 1 or 2 MHZ per horizontal division. 
What you will most likely find is your IF signal (over at 8.9MHZ) is nearly as strong as your desired signal, While you are transmitting...

While you are looking at the analyzer (and in transmit mode) rock the preselector control back and fourth. Look to see if the frequency of any of the signals on the display are changing in frequency, The desired and IF frequencies will change in amplitude... Rotate the preselector and see where the desired signal peaks...
See if that is where the manual says it is suppose to be at or is it further counterclockwise from that position.

Because the IF 9.8MHZ is close to the desired frequency  7 MHZ it is easy to misalign.

There ya GO. 


 

[drforbin]

Hi..thanxs....I tested for dc before using SA and I coupled with a high cap thus eliminating dc. Anyway. I think your input is great and thank you so much.
As I said before I really didn't remember you asking for a specific test.

I'd be happy to do what you ask for.
I did wish to bring you all up to date.

I just got finished performing my test.
I isolated the driver/pa from the rest of the unit and used my new (old) wavetek 288 to inject a 7Mhz signal into the driver.
I had my DMM connected to ALC output.

Using the signal from the wavetek or the signal from the radio I GET the same ALC offset.
This to me clearly means the signal leading to the driver/final is fine.
What do you all think?

If you still wish me to conduct any test let me know and I shall do it.

[voltz]

Firstly, congratulations on getting a fine function generator like that. Nice piece of kit.

If you're taking requests for tests, i would like to see the RF Out phono jack socket at the back with SA please. Tuned on 40m. ALC on. +- sweep.
Thanks.

[drforbin]

Ok voltz...I'll do it.

What do you think about my test?
any comments?

[voltz]

I have to agree, it removes any chance of IF being injected, or any other unwanted frequencies for that matter. But the fault remains a mystery at this time.
I am hoping the RF Out jack will show us something interesting. It is sensing the exact same point the ALC does. So we should see something! Its hard to guess the sweep range this time, if this is resonance, it could be anywhere.. So sweep away and have a good look around, up and down, then close in on 7. Thanks for doing these tests.