advertisement


Radford STA15 Restoration

Do the handles bolt through the back plates, i.e. with a nut either side?
If so, remove back plate and nuts, clamp one threaded leg in soft jaws in a vice, place a tube over the other leg and twist back to shape. Must be a fishie local to you who can do this? If not, I can do them for you.
 
Are they twisted or bent?
Remove the nuts and 'trim plate' and lay them on a flat surface.
If they are bent one vertical will be flatter than the other or one of the threaded sections will be askew. You need to work on that vertical.
If they are twisted then you should be able to twist them straight again with man handling.
Don't aim for perfection - none of them is to start with.
 
Loosen the screws, put the flat part into a door jamb and apply a little pressure in the correct direction, should be easy enough.
 
The handles are twisted rather than bent. As if someone has taken each upright and twisted them in different directions. Mr a.f, I may take you up on your offer if the door jam doesn't do the job.

A little more progress, wiring done...
2022-05-24_11-34-49 by Garf Arf, on Flickr

I'll have to spend a bit of time checking the wiring and dressing it before I apply some power on a dim bulb tester.

And the 8uF preamp PSU cap is too big to mount under the PCB so I had to build a bit of a heath-robinson affair on top of the PCB.

I don't think I have ever stripped and tinned so many bits of wire or de-soldered and cleaned so many connections. Let's hope there are no shorts. :)

And of course I have the external bits of the amp to clean up yet.

Bedtime...
 
Last edited:
Did all the usual checks before dim bulb power on; checked for shorts between the mains and earth, HT and gnd, heater cct and gnd,5V cct and gnd and general checking that all gnds are connected. Quick power on with no valves to check the mains transformer outputs and all looked good. HT, heater ccts and 5V lines all quite high as there was no load on them.

Installed the GZ34 (strangely the amp came with a GZ32 installed!). Power back on with the bulb in the line. This time I measured about 450VDC over the PSU caps; high but at least the rectifier is working. Took the buib out of the main line, power back on and still 450vDC over the PSU caps, looking promising.

Installed the valves on one channel, added the bulb back into the mains and powered back on. Well, no smoke and everything came up fine. Voltages a little high, but I would expect that with no load from the other channel. Removed the bulb, re-powered up and still no smoke :) Voltages across the cathode resistors where about 28V on the first channel.

Installed the valves on the second channel, bulb back in the cct and power on. Again, no smoke, but the voltages are struggling now with the strangle hold of the bulb. Time for full power...

And, well, everything came up, the HT voltages around the boards were pretty close to those on the cct diagram. The only differences I could find were across the cathode resistors. Now, the amp came with two different sets of EL34, both sets Mullard, but one from the mid 60s and one from the late 60s. I haven't double checked which set was which, but with one set the voltages across the 470R resistor were about 22VDC and about 26VDC across the other set's 470R resistor. I would assume then that that indicates that one set of valves is a little more tired than the other set? I will swap the valves over tonight and confirm that the voltage difference moves to the other channel, just to be sure it is the valve and not the boards.

I did manage to run the amps, in to a dummy load for about an hour without anything disastrous happening, or bursting into flames.

I ran out of time to actually run some audio through them, a job for tonight.
 
Thank you Will. Fingers crossed that it works.

Do you know the size of the screws that hold on the base plate at all? My amp came with two of them missing.
These things add character. Mine has been missing a cover screw for the whole of my ownership (35 years.)

(Although if I’m honest the reason for not replacing it is that I can’t be bothered to track down a suitable clip thingy for it to screw into, which is also missing.)
 
These things add character. Mine has been missing a cover screw for the whole of my ownership (35 years.)

(Although if I’m honest the reason for not replacing it is that I can’t be bothered to track down a suitable clip thingy for it to screw into, which is also missing.)
Is that why my wife calls me a character cos I've got a screw lose? :)
 
...I've got a screw loose... :)

You and me both. After posting about the screw I remembered I hadn’t lost it, I was simply missing the clip. I had wrapped the screw in red insulating tape and put it somewhere for safe keeping. I then spent half an hour searching through the more obvious boxes and drawers. I gave up, then over a cup of coffee thought ‘hang on, something in my memory tells me it is close to the amp...’ Aha!

52101800792_57357f351e.jpg
 
You and me both. After posting about the screw I remembered I hadn’t lost it, I was simply missing the clip. I had wrapped the screw in red insulating tape and put it somewhere for safe keeping. I then spent half an hour searching through the more obvious boxes and drawers. I gave up, then over a cup of coffee thought ‘hang on, something in my memory tells me it is close to the amp...’ Aha!

52101800792_57357f351e.jpg
A great example of why us old chaps need regular stimulants in order to be a benefit to society. :)
 
Well, it isn't all fun and games it seems. Before plugging the STA15 into my 57s I thought I'd plug it in to an oscilloscope, and I'm glad I did as one channel was misbehaving, going into oscillation with the amplitude of the oscillation quickly shooting off the scale. I swapped the valves over between the channels and the oscillation stayed with the channel. Hmmmm. I checked and double checked all my wiring connections both for good solder joints and connecting the right bit to the right bit - at least I had one channel that was working to compare to. But I couldn't find the mistake. I even buzzed out all the wires, all the transformer resistances, double checked for solder bridges on every connection, everything that I could think of.

Strangely, the only different in voltages I could find was between the channels was that over cathode resistors; one channel as about 25.5VDC and the other 21VDC.

I was a little stumped, I mean the wiring is not that complex. So what's the best thing to do in those situations? Go to bed and sleep on it, and I followed my advice to the letter.

Thought I'd have another go today.

I took the misbehaving PCB out by unsoldering the top output connections and the HT connection, but leaving all the connections to the valves in place. When I double checked all the soldering on the PCB I couldn't see anything out of line, all the resistances between the two PCBs where the same, nothing leapt out at me. I did rework one connection between the PCB and the ECF82, but only because the routing of the wire didn't quite match my OCD. Without much hope I re-soldered the top connections and the HT, screwed the PCB in and hooked up the sig gen and oscilloscope and the bloody thing worked fine. Checked all the three speaker taps, still worked fine. Poked around the PCB and the valves with a chopstick, couldn't upset it. So, I am in the dark, but it works...yeah :)

I took the opportunity to run the amp up to signal chipping to measure the output power. I don't have the ability to measure THD so my measurements are kinda crude.

Previously ill channel (Yellow)

SDS00033 by Garf Arf, on Flickr

SDS00034 by Garf Arf, on Flickr

SDS00040 by Garf Arf, on Flickr

Working all the time Channel (Purple)

SDS00037 by Garf Arf, on Flickr

SDS00038 by Garf Arf, on Flickr

SDS00039 by Garf Arf, on Flickr

Both channels are peaking out at about 13.2VDCrms in to an 8R dummy load which gives a power per channel of (V^2/R) 21W. I would imagine the THD is quite high at that output.

The square wave of the yellow trace does have a very slight "ring" to it but nothing bad. In fact I was expecting to see a much more rounded set of square waves, much like the ones the Quad IIs generate. But the STA15's are very straight up and down.

So, time to put the handles on, the baseplate on and slide the cover on. After that I'm gonna check it again on the bench, just to build some trust between us both.

Tomorrow I see speakers in the mix...

PS. I've ordered a pair of NOS Telefunken ECF82s

PPS. Is there somewhere I could have these Mullard EL34s tested? I would like to know how "fresh" they are.
 
I got to this stage a few years ago, albeit the (really just a check-over) work had been done by someone else. My STA15 had been lying under a bed for twenty years while I listened to a succession of Naims and Quads. I got it home, left it until the next day...

...then I plugged it and the SC22 into a pair of BC1s in a freezing cold spare room just to see if everything was working correctly. Three hours later I was still there and knew I had found my amp for life. Then, inspired, I went Tannoy hunting, and found my speakers for life too. You have been warned.
 
Last edited:
I got to this stage a few years ago, albeit the (really just a check-over) work had been done by someone else. My STA15 had been lying under a bed for twenty years while I listened to a succession of Naims and Quads. I got it home, left it until the next day...

...then I plugged it and the SC22 into a pair of BC1s in a freezing cold spare room just to see if everything was working correctly. Three hours later I was still there and knew I had found my amp for life. Then, inspired, I went Tannoy hunting, and found my speakers for life too. You have been warned.
I hear your warning, and have heeded it. Off shopping this morning, left the amp on the bench, going to try it later this afternoon after I've done my chores.

I'd like to know the background of how your Radford spent 20 years under a bed. Was it a talisman of some sort, or banished it to the darkness?
 
Is there somewhere I could have these Mullard EL34s tested? I would like to know how "fresh" they are.

You may find a guitar shop with a service area has a valve tester. I’ve got a pretty basic Orange (guitar amp brand) one which does EL34s, but it wouldn’t be worth risking shipping them to me to test. I’m not sure how much value it adds anyway. If they are working fine, not drawing unusual current, red-plating, needing odd bias voltage etc I’d expect them to be fine. The fact you are getting good clean waveforms out the back suggests they are fine. I should really try that with my Leaks at some point as I’ve never stuck a scope on the output (I do have one).
 
You do not say if the EL34 cathode voltages are the same or better since you stopped the oscillation? Personally I would pop a set of JJ EL34s in, or what ever make you prefer, and keep the Mullards safe.
I normally 'pair them' in the amp by measuring their individual voltages, then put the closest values back as pairs. ( Say one is 24 volts, two are 25 volts and one is 26 volts, put a 24 and 25 together, then put the 25 and 26 together.)
I have a pair of Mullard ECF82s that are noisy, they work fine, but whistle when warming up and add a slight 'fizz' to the between tracks quiet sections. I have since gone to Canadian or US 6U8s they have better (closer to the diagram) voltage readings and are quiet...
 
You do not say if the EL34 cathode voltages are the same or better since you stopped the oscillation? Personally I would pop a set of JJ EL34s in, or what ever make you prefer, and keep the Mullards safe.
I normally 'pair them' in the amp by measuring their individual voltages, then put the closest values back as pairs. ( Say one is 24 volts, two are 25 volts and one is 26 volts, put a 24 and 25 together, then put the 25 and 26 together.)
I have a pair of Mullard ECF82s that are noisy, they work fine, but whistle when warming up and add a slight 'fizz' to the between tracks quiet sections. I have since gone to Canadian or US 6U8s they have better (closer to the diagram) voltage readings and are quiet...
Al, the voltages across the cathode resistors on both channels have all settled between 26.6-25.5VDC, no longer around 21VDC for the oscillating channel.

In fact the grouping will be interesting as each pair of EL34s are from different years, one pair with single getters and the other pair with double getters.

I'm going to listen to the Mullards for a while, but I do take your point about the JJs. I believe Radford Revivals use JJs in their new amplifiers.
 
I'd like to know the background of how your Radford spent 20 years under a bed. Was it a talisman of some sort, or banished it to the darkness?
It’s a long (around 20 years, I’m guessing) and involved story. It involves familiarity breeding contempt, thinking I could do better (fuelled by other folk’s writings), wondering why the new didn’t sound as good to me as it did to them, endlessly adding, subtracting and tweaking to solve the puzzle (passing through Isobariks, ESLs, ESLs+subs, Naim amps, Quad amps, home brew valve amps, home brew SS amps, DSP et al) and getting totally bewildered - not realising I was in fact slowly inching further away from what I really wanted, and that the (correct) whole is greater than the sum of the parts. Finally I decided to flog the lot and go back to square one. Eureka!
 
Last edited:


advertisement


Back
Top