Dynatron Nomad - impressive!

[llama]

I've just opened up a Dynatron Nomad which I've had for years, probably from a boot sale so would have been a couple of quid at the most.

I see from R&TVS (1960-'61) that my Nomad is a TP11 and the very similar TP12 is a table-top set.

The constuction of the Nomad is very impressive indeed and I hope the output is up to the standard of construction. Here are some of the high-end features I've noticed:

Separate, very long, ferrite rods for medium and long waves

Fixing self tappers for the chassis go into proper spring steel clips

Speaker leads just unclip

Really solid looking volume control / 2-pole switch (needed for the twin PP9s)

Outstanding tuning capacitor incorporating slow motion drive and anti-backlash gearing

PCB looks to be good quality.

My plans are to get this running and then fit a BlueTooth module for my grand-daughter to use as output for her tablet. The Plessey electrolytics are likely (I'm told) to need replacement so that'll be the first job. The Hunts little black capacitors will only get changed if they measure faulty. I can't say I like the resistors used throughout - metal caps pressed onto the ends of (I presume carbon composition) rods. Again, I'll only change them if they turn out to be faulty.

Observations from others welcomed.
Graham

[llama]

I have changed the 7 "red/black" Plessey electrolytics. No problem with the 100uF ones but I don't have any new 8uF ones so had to fit the two 10uF caps I've got and a 22uF to replace the last 8uF. I put the 22 in the signal path and the 8uF on the AGC (I presume) line is now a 10.

Much to my relief (having read previous posts) the speaker works very well. In fact the set seems very sensitive on Medium Wave. The fact that some stations crack up when tuned to central carrier makes me wonder if the AGC is working. However, I've only got PP3s at the mo so maybe they are flagging. I'll probably use multi-AAA holders ultimately.

The wavechange switch is jammed. I'll need to get the chassis out again to fault-find that. Even though LW won't get used much if at all I feel I should fix it if poss.

In line with other people's posts, there is significant hiss on zero volume setting. I might try an OC42 in the audio amp.
Graham

[G6Tanuki]

The "stick" resistors can be a cause of noise - the pressed brass end-caps go brittle and crack along the tubular bit, meaning they no longer exert much pressure on the carbon-stick and the poor contact can lead to all sorts of fizzles.

I'd check the ones in the first stage of the audio-amp before you start replacing transistors.

[Ed_Dinning]

Hi Graham, I have plenty NOS resistors like that if originality is important.

Never used, never solderd so may be OK. Mail me if you want any

Ed

[Electronpusher0]

Funnily enough I am working on a Nomad at the moment also.
I bought it at a car boot some time ago but only got it on the bench today.
As you say, very well built.

Mine has a problem with the local oscillator. If I start with the variable capacitor vanes fully out it starts, it continues to run (and tunes stations) until the vanes are about 2/3rds mshed when it stops.
However, to get it to start again requires fully opening the vanes again.
The tuning cap vanes are not shorting out at any point.

It is obvious that this is not a new fault, someone has changed several caps around the oscillator, the tuning cap vanes had an oily residue on them, looks like some has used switch cleaner on them. I have flushed away the residue with solvent.

I have changed the resistors around the oscillator transistor and the transistor itself but the fault is the same. I have left off for tonight and will resume tomorrow.

I also have found a lot of hiss on zero volume setting, I have tried changing the first AF transistor with a different type it made no difference. I have parked that issue until the oscillator is fixed.

When I eventually get it going I also intend adding a bluetooth module (in a way that could easily be removed and the set restored to its original condition).

Peter

[Electronpusher0]

I have just returned to the bench and found the fault with the local oscillator.
It was X1 which forms part of the AGC system.
I replaced it with a NOS OA90.
Now to tackle the hiss.

Peter

[Electronpusher0]

I think I have solved the excessive hiss.

To isolate the audio stage I temporarily lifted C31, the coupling capacitor from the wiper of the volume control. I measured the noise on the loudspeaker with a scope, it was approx 50mV p-p.

I initially concentrated on the first audio amp, TR4, as I expected the source of the noise to be in the front end before any gain was applied.
I changed all the resistors around TR4, they had all drifted well off the correct value and broke as I desoldered them. This had no effect!
I had already changed TR4 to an AC126 but there was no change so I changed it again to a silicon PNP device. I changed the bias by changing R16 from 180K to 82K, this restored the voltage on the collector to the correct value.
Unfortunately this also did not work.

I next lifted the coupling capacitor between the collector of TR4 and the base of TR5, C35. This removed any possible noise contribution from the first AF stage. Again, no difference.

Finally I changed the driver transistor, TR5, to a silicon PNP device. I made no changes to the bias. The base of TR5 is held close to 0V by R22, changing to a silicon device simply changed the emitter voltage from 0.2V to 0.6V, this had minimal effect on the collector current which is set by the voltage across R23/R24, this voltage changed by 0.4V in 9V so minimal effect.

When I switched back on I had to double check the amplifier was actually working so quiet was it.

Restoring the lifted coupling capacitors resulted in a working radio with no hiss from the audio stage.

I will now replace the electrolytics as a precaution, clean up the case and add the bluetooth.

Peter

[llama]

Emboldened by your findings and subsequent modification, I went straight for TR5 and swapped it for an MPS6516 - simply because it has EBC moulded into the casing. Unfortunately there was no improvement in the hiss level so I'm going to have to be more scientific about it.

I freed up the wavechange switch by dripping some light oil down the shaft from the top. Working the switch a few times and the back of the knob broke. My bad. It should super-glue together.
Graham

[Radio Wrangler]

I notice that it doesn't have any decoupling capacitors on the supply rails after the power switches and before the output transistors. I think this set may be a bit intolerant of rising ESR as batteries run down.

David

[Electronpusher0]

Originally Posted by Radio Wrangler

I notice that it doesn't have any decoupling capacitors on the supply rails after the power switches and before the output transistors. I think this set may be a bit intolerant of rising ESR as batteries run down.

David

Agreed, I have added a 2200uF capacitor across the supply after the on/off switch.

Peter

[Electronpusher0]

Originally Posted by llama

Emboldened by your findings and subsequent modification, I went straight for TR5 and swapped it for an MPS6516 - simply because it has EBC moulded into the casing. Unfortunately there was no improvement in the hiss level so I'm going to have to be more scientific about it.

Graham

Thinking about it I also left the change of TR4 to a silcon device in place so you may need to change both devices.
Try lifting one end of C35 to separate the 2 stages, if the hiss goes (with TR5 changed) then you will also need to change Tr4.

Peter

[Radio Wrangler]

Because it uses the centre point between the batteries for the cold end of the speaker, it really needs two decouplers.

David

[llama]

Progress.
Having replaced a couple of the brass-cap resistors with no effect, I tacked a 150k resistor across the 180k base bias resistor of TR4 (R16 in the thumbnail in post #7 - not numbered in the RTVS book) that equates to the 82k adopted by Peter and fitted another MPS6516, Si, for TR4.

Nice and quiet now, though even at minimum volume there is a tiny bit of programme coming through. Noise well below that though. It's one of the potential effects of having the on-off switch on the volume control. I might reduce the overall gain to minimise that.

Now to look for a pair of high value electrolytics as power supply decouplers.
Graham

[llama]

I'm stuck and would like a bit of help, please.
Firstly though I should add that I've wired-in 2200uF electrolytics on the two rails out of the on-off switch but they haven't affected my problem which is ineffective AGC. My measurements tell me I don't really know how AGC is supposed to work in transistor radios.

I've measured the Ge diode that's deep inside the RF stages (but actually tacked on the back of the PCB). It measures .29 v fwd on my diode test and high resistance in reverse. I presume the main signal diode is OK - it demodulates the signal and produces DC for the AGC. Measuring the AGC line with my AVO 8 gives 1.9 v when tuned to a strong signal and 2.7 v when as close as I can get to no signal.

I'd appreciate help in decoding what these measurements might indicate. Is that a typical voltage range for a transistor set AGC line, for instance?
Graham

[ms660]

2.7 volts on the AGC line with no signal, it could be something wrong with the detector diode circuit or the bottom half of the potential divider to ground, eg: R14, R15.

Lawrence.

[Electronpusher0]

I forgot to add that I also had to change the AGC diode on my Nomad.

Peter

[llama]

Hello Peter - you mentioned changing X1 in post #6. That's the diode I measured and seems OK.
Hello Lawrence - the potential divider seems to be 56k to HT which biases TR2 via IFT1, then 8.2k to the detector diode which returns to ground via R14 and the volume control. I wonder what a typical voltage range would be.

Thanks for the input chaps. Does the AGC just work by starving TR2 of bias under strong signal conditions?

I'm wondering if the .04uF C21 could be leaky and upsetting AGC by shorting it nearer to ground via R9. But I think that might have the opposite effect. Like I say, I'm struggling to understand the AGC mechanism - and why there are 2 diodes.
Graham

[ms660]

Originally Posted by llama

Like I say, I'm struggling to understand the AGC mechanism - and why there are 2 diodes

The 8.2k is connected to ground via the detector diode in series with the secondary of the transformer and in parallel with R14, R15.

Don't forget that the top supply rail is -ve, increasing the signal increases the +ve output from the signal detector X2, that in effect reduces the -ve base bias to TR2, thus reducing the gain, over a certain signal strength the voltage on the collector of TR2 will rise above X1's threshold voltage due to the AGC action provided by X2, in other words the voltage on X1's cathode becomes approx 0.2v (ish) higher than the voltage on its anode, that causes X1 to conduct, in doing so it connects R4 and R8 (that are in series) across part of the primary of the 1st IF transformer, the damping effect of that thus reducing the gain even further without having to bias TR3 close to cut off.

Lawrence.

[llama]

Oo - a lot to digest there.
Yes, my quoted AGC voltages were both negative wrt ground. So with no signal the (neg) 2.7v biases TR2 more strongly vs (neg) 1.9v biasing it less with strong signals. I guess faulty resistors in any of the DC paths around X1, X2 and TR2 would upset the rather delicate balance between the right amount of gain and too much gain. I don't like a blanket-replace but changing them is probably no more work than lifting a leg to measure.
Thanks again.
Graham

[llama]

Seeing as R6, 8.2k, feeds the whole of the AGC network, I found that under the detector stage screening can. Out of circuit it measured a bit high and a new 8.2k went in. That made the AGC range 1.7 v to 2.5 v but not audibly different from how it was.

Next to measure/change was R5, 56k. This measured about 70k and a new replacement brought about a radical change in the AGC range - 2.4 v strong signal to 3.2 v no signal. However, the result was audibly worse.

There are still 2 or 3 resistors affecting the operating points in that area so I'll have to attend to those to get to the bottom of this. As stated earlier, the diode above TR1 measures good and the detector diode detects so that should be fine.
I wish I could be more methodical about this but it all seems to be so interdependent.
Graham