UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Specific Vintage Equipment > Vintage Audio (record players, hi-fi etc)

Notices

Vintage Audio (record players, hi-fi etc) Amplifiers, speakers, gramophones and other audio equipment.

Closed Thread
 
Thread Tools
Old 5th Mar 2017, 1:04 pm   #1
Valvepower
Octode
 
Valvepower's Avatar
 
Join Date: Aug 2008
Location: Rayleigh near Southend-On-Sea, Essex, UK.
Posts: 1,852
Default Spot the UA15 Black Box 1004 part!

Hi,

Just started work on the restoration of a PYE transistor Black Box and I’m trying to identify a part (picture attached), which was Sellotaped to the turntable, so it figures at some time in its past it was removed? If it helps the cartridge has been replaced with a GP91-3SC and I’m wondering if it was removed from the tone arm wiring when the cartridge was replaced?

The amplifier looks in good shape having not been ‘got-at’ plus the output transistors read ok for no shorts. I’m going to re-cap it to be safe.

This is one of the "Black" Black Boxes in piano black finish.

Regards
Terry
Attached Thumbnails
Click image for larger version

Name:	UA15_Thing_01.jpg
Views:	239
Size:	44.5 KB
ID:	138619   Click image for larger version

Name:	UA15_Thing_02.jpg
Views:	219
Size:	48.1 KB
ID:	138620  
Valvepower is offline  
Old 5th Mar 2017, 1:15 pm   #2
tracypaper
Octode
 
tracypaper's Avatar
 
Join Date: Aug 2010
Location: Beeston, Leeds, West Yorkshire, UK.
Posts: 1,234
Default Re: Spot the UA15 Black Box 1004 part!

That is from the top of the tower on the deck where the record size selector is located, it it from the cap and is there to clip the cap in place, originally it was pushed onto a plastic pin (part of the cap) and then the pin was melted to hold it in place, they are often missing as the clip pops off when removing the cap to access the size selector.
Gaz.
tracypaper is offline  
Old 5th Mar 2017, 2:55 pm   #3
Edward Huggins
Dekatron
 
Join Date: Sep 2015
Location: Southwold, Suffolk, UK.
Posts: 8,302
Default Re: Spot the UA15 Black Box 1004 part!

That Acos cartridge is the wrong one for this player. It's a high-output type (so worth quite a bit in its own right) but with 700mv output it will overload the AF stages of the early germanium amplifier and cause much distortion.
As to being Piano-Black, I'm not aware that we ever produced the 1004 in Black - not even as option or for export. Very strange. Home re-spray?
__________________
Edward.
Edward Huggins is offline  
Old 5th Mar 2017, 3:42 pm   #4
Hartley118
Nonode
 
Hartley118's Avatar
 
Join Date: Aug 2015
Location: Cambridge, Cambs. UK.
Posts: 2,196
Default Re: Spot the UA15 Black Box 1004 part!

Did it once have the sought-after 'Chinoiserie' painted designs on it I wonder?

Edward - were they hand painted?

Martin
__________________
BVWS Member
Hartley118 is online now  
Old 5th Mar 2017, 4:58 pm   #5
Valvepower
Octode
 
Valvepower's Avatar
 
Join Date: Aug 2008
Location: Rayleigh near Southend-On-Sea, Essex, UK.
Posts: 1,852
Default Re: Spot the UA15 Black Box 1004 part!

Hello,

Gaz, thanks for identifying the part, it does make sense as I can see in the past where some tape had been put over the top of the record size selector.

Edward, many thanks for that, as I must admit to being unsure of the suitability of the cartridge because as far I could make out the Ronette TX88 output level is 160mV, whereas the GP91-3SC is 630mV. Also, as well as overloading the first stage I suppose the high output cartridge would have upset the series 'loading' volume control on the input stage of the first transistor VT1.

I have a GP91-1SC in the workshop, which gong by the data on the carton appears to have an output of 200mV, which is nearer that of the TX88.

It’s defiantly a factory finish as on disassembling the unit there are no signs of overspray, also on the inside, you can see the sings of it being done by the factory.

I know the transistor Black Box doesn’t have the kudos of the valve ones, but for me it represents a piece of history with the early use of the Quasi complementary output stage. I suppose the transistor circuit would have been quite a departure PYE at the time. Would I be right in thinking this was the work of T. D. Towers? Based on the Tobey Dinsdale circuit.

Martin, can’t see any signs of a 'Chinoiserie' painted design.

Regards
Terry
Valvepower is offline  
Old 5th Mar 2017, 5:07 pm   #6
Edward Huggins
Dekatron
 
Join Date: Sep 2015
Location: Southwold, Suffolk, UK.
Posts: 8,302
Default Re: Spot the UA15 Black Box 1004 part!

Yes, the Acos GP91-1SC will be ideal - but do save that GP91-3SC, rare as hen's teeth and valuable! Ensure you re-balance the UA15 arm.
The transistor circuit was advanced as it's design started as early as late 1962. It used the transistors made in Pye's, then new, Factory in Newmarket.
I certainly remember the Dinsdale Circuit and as to whether Mr. Towers lifted some of it, I couldn't possibly comment.....!

The true "Black Box" cabinet was only ever used on the very first Pye Black Boxes (EL42s) from 1954 to early 1956. The Chinoiserie decoration in the greater part, and to the best of my knowledge, were Transfers. Some got distorted upon application. By mid-1956, when the then "new" BSR UA8 4 speed decks were being fitted, no more were produced and the pre-made cabinet stock from Cabinet Industries' factory were used up.
__________________
Edward.
Edward Huggins is offline  
Old 6th Mar 2017, 9:29 pm   #7
Valvepower
Octode
 
Valvepower's Avatar
 
Join Date: Aug 2008
Location: Rayleigh near Southend-On-Sea, Essex, UK.
Posts: 1,852
Default Re: Spot the UA15 Black Box 1004 part!

Hello Edward,

Thanks for the reply.

Thankfully all the transistors appear ok so a slice of Newmarket transistor technology has been saved. Thankfully the seller had refrained from powering it up and thus avoiding the risk of damaging the output transistors.

Using the GP91-1SC I may have to fettle R3, which is in series with the cartridge output. According to the service sheet this resistor was selected during manufacture.

I’ll put the GP91-3SC aside for the next single stage Dansette type record player that comes my way.

I noticed a thread on the forum about the preferred version PYE Black Box, although this will not have the ‘sound’ of the valve ones and is seen as the poor relative, it has its place as the first transistor Black Box.

To improve the sound and elevate the Quasi complementary distortion I suppose I could fit a Baxandall diode to the bottom NPN/PNP pair and re-bias the output stage, but I could be tempting fate with the risk of popping the output transistors if I were to make a mistake plus would it really gain me anything.

Regards
Terry
Valvepower is offline  
Old 7th Mar 2017, 10:44 am   #8
Edward Huggins
Dekatron
 
Join Date: Sep 2015
Location: Southwold, Suffolk, UK.
Posts: 8,302
Default Re: Spot the UA15 Black Box 1004 part!

I would leave the bias and the overall circuit as is. It's already as good as it can get. There are near equivalents of the Nkt Germaniums if ever needed though. And that saved Acos will be ideal in a single-stage player.
__________________
Edward.
Edward Huggins is offline  
Old 7th Mar 2017, 11:15 pm   #9
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,943
Default Re: Spot the UA15 Black Box 1004 part!

Re the use of a Baxandall diode in a germanium Lin-type quasi-complementary output stage, I am not aware that it was ever done commercially. In part that could be been the result of chronology, in that by the time the Baxandall diode arrived (1969), most amplifier makers were using silicon output devices. Still, it might have been adopted by the late users of germanium, and some of the Bryan amplifiers of the late 1960s/early 1970s could be places to look for it.

On the other hand, it could be said that it was the shift from germanium to silicon that drove the need to develop much more symmetrical versions of the Lin-type output, in that the ill-effects of the inherent asymmetry of the original version were much worse with silicon than with germanium transistors. Hence the Quad triples (1967), the Shaw diode (1969) and the Baxandall diode (1969), the last-mentioned becoming something of a norm. In qualitative terms, my understanding is that the fast switching of silicon transistors resulted in transfer curves with a much tighter turn-in at the switch-on/off point than with germanium transistors, and which reached their relatively straight sections very early on. Also, the conjugate pair I think had an even tighter turn in than the Darlington pair. Coupled with its inherent greater steepness, this resulted in a net transfer curve that had a marked kink (possibly a non-differentiable point) at the crossover point. On the other hand, the much slower curvatures associated with germanium transistors meant that the net transfer curve probably had a rather gentle change of slope whose curvature extended quite a way on both sides of the crossover point.

I understand that those makers who retained germanium output devices in the face of the general change to silicon sometimes claimed lower crossover distortion as their reason for so doing. So by the late 1960s, the options for minimizing crossover distortion appeared to be:

1. Use the Lin circuit with silicon transistors but pile on the NFB and look the other way.
2. Use a transformer drive.
3. Use a Class A output.
4. Use silicon transistors in a modified Lin circuit, such as with the Baxandall diode.
5. Use a fully-complementary circuit
6. Use the Lin circuit with germanium output transistors.

Whether (6) alone was enough or whether it would potentially be advantageous to add the Baxandall diode I don’t know. But I think that it is a question that ideally would need to be answered before contemplating the modification of an existing circuit, and even then, results might vary situationally. I’d guess that the Dinsdale circuit might be a good basis for experimentation.

A pertinent observation was that Bailey’s 1966 design (Wireless World (WW) 1966 November) was offered with both germanium and silicon output transistors, the latter where greater robustness was required. Germanium was suggested for domestic use with the comment: “In fact germanium transistors usually give far lower distortion due to their better linearity”. This was a transformer-driven design, so it was not so much an issue of output stage asymmetry, but rather related to the inherent properties of the base semiconductor materials. Combine this with the fact that germanium transistors tend ameliorate the deleterious effects of Lin circuit asymmetry, and it seems possible, even probably that no further refinement is really needed, particularly at say the “mid-fi” level.

By the way, Bailey’s reason for choosing transformer drive in 1966 were more clearly articulated in the preamble to his 1968 fully-complementary design (WW 1968 May), in which output stage symmetry was discussed. Nevertheless, the assertion that the fully complementary circuit was inherently symmetrical was challenged by others, notably Julian Vereker of Naim.

To conclude, I’d suggest that the available evidence points to a “leave well alone” approach in the case of the Pye 1004, as suggested by Edward, unless further research indicates demonstrates that there is definitely a prize to be had that is worth the hassle of getting it. The first step might be to ascertain whether late germanium users such as Bryan ever did use the Baxandall diode.


Cheers,
Synchrodyne is offline  
Old 9th Mar 2017, 10:15 pm   #10
Valvepower
Octode
 
Valvepower's Avatar
 
Join Date: Aug 2008
Location: Rayleigh near Southend-On-Sea, Essex, UK.
Posts: 1,852
Default Re: Spot the UA15 Black Box 1004 part!

Hi Synchrodyne,

Many thanks for the great input about quasi complementary and Baxandall diode. I’ve decided to leave it well alone as there is a risk of damaging the output transistors if things should go awry, also for the sake of authenticity
it’s prudent to leave as is.

I found the idea that the shift from germanium to silicon exaggerated the ill-effects Lin-type output, where it’s worse with silicon than with germanium transistors. I also found interesting the idea that the ‘slower’ germanium were more forgiving and lessened the sonic effects of the quasi complementary asymmetry.

I feel there is something in using a driver transformer – although not ideal it does offer some advantages.

Regards
Terry
Valvepower is offline  
Old 26th Mar 2017, 7:17 pm   #11
Valvepower
Octode
 
Valvepower's Avatar
 
Join Date: Aug 2008
Location: Rayleigh near Southend-On-Sea, Essex, UK.
Posts: 1,852
Default Re: Spot the UA15 Black Box 1004 part!

Hello,

Very quick update on the Transistor Black Box… I’ve got it running out of the cabinet with a bench speaker with a reasonable amount of volume, although not as much as I’d thought there would be, but it’s enough.

After re-capping, I very slowly powered it up using an Variac and quite quickly, with not a lot of mains volts, I realized things weren’t right when the centre point voltage (neg end of C12) hardly rose above ground and the quiescent current was far more than the 20mA. I did another check of the transistors looking for a short circuit transistor, but I couldn’t find anything obvious.

Carefully looking at the fault conditions I had my suspicions about VT3 and VT4 and by substituting VT3 and then VT4 I found VT4 (NPN driver) was failing under load. I replaced this with a 2N1308 I had in the workshop and it came to life! It’s worth noting the unit was fitted with a NKT773 and not a NKT751 as per the service sheet.

Set the mid point pre-set and connected a sig gen, scope and load and I was getting a healthy signal with a symmetrical clip.

The tone pot needed a clean, but I replaced the volume pot with a 100K Citec Dual pot where I replaced one of the tracks with a 220K track from a donor pot.

The UA15 was a bit of a dog’s dinner as the motor bearings needed cleaning and re-soaking with oil. Re-greased the auto mechanism and centre bearing and cleaned the auto shut down parts. The auto shutdown is a tad erratic, working most of the time, but every so often I must manually shut it down!

I fitted a GP91-1 to the UA15..... I’d checked the level of this cartridge and found it to be giving a healthy 170mV with a 1.2cm/sec 1K test LP, and going by the data I found on the TX88 I suppose this is as near as I’m going to get to the TX88 level.

I must say the first stage (VT1) with the series loudness control and EQ network C14, R16 & C3 is a clever 'ole circuit. This with the Dinsdale “like” power amplifier makes it an interesting piece of early transistor technology.

Next thing is to clean the cabinet and put the whole kit and caboodle back together in the cabinet.

Regards
Terry
Valvepower is offline  
Old 26th Mar 2017, 9:46 pm   #12
Edward Huggins
Dekatron
 
Join Date: Sep 2015
Location: Southwold, Suffolk, UK.
Posts: 8,302
Default Re: Spot the UA15 Black Box 1004 part!

Don't be content with a low output. This amp should give a very clean 4 watts. The Acos GP91-1 has measured well and should provide an adequte load to the amplifier. Is the lack of gain in the AF/Driver stage? The auto shut off on the UA15 is very common foible, in part due to a lack of torque. I would not worry about it, just manually switch it to Off.
__________________
Edward.
Edward Huggins is offline  
Old 29th Mar 2017, 7:13 pm   #13
Valvepower
Octode
 
Valvepower's Avatar
 
Join Date: Aug 2008
Location: Rayleigh near Southend-On-Sea, Essex, UK.
Posts: 1,852
Default Re: Spot the UA15 Black Box 1004 part!

Hello,

I’ve checked the output power and it measured 5.6 Watts into a 7.5 Ohm load at the onset of clip, but using the Acos cartridge I feel I must turn the volume further up than I’d expected to get higher levels of volume – if you see what I mean.

With the volume at maximum and the tone set for no HF cut the input sensitivity for 5.6 watts is 600mV, which makes the gain of the complete amplifier 20.4dB.

From the circuit, I calculated the gain of the complete amplifier and this came in at 22.6dB so I felt the measured gain of 20.4dB to be satisfactory given the 10% resistors and the assumptions I made in the calculating the gain.

The power amplifier was measuring as I’d expect and to make sure the first stage was ok I replaced VT1 for a 2SC440, which made little difference. I was kind ‘a surprised the first stage has a gain of -17.59dB at 1KHz so I double checked this and it appears ok compared to the calculated gain of -16.71dB

I’m wondering if the Acos has a slightly different source impedance to that of the TX-88, which acts differently with the lower input impedance of the preamplifier and not giving as much level as the TX-88? Or PYE restricted the maximum level to eliminate feedback and other similar problems?

But… after a bit of thought and another look at the original Dinsdale preamplifier circuit from 1961 I decided to reduce R3 to 100K and this appears to give me enough level.

Forgot to say, all the DC voltages measure ok compared to the voltage chart, given the resistors are 10%

It does a fairly good job of playing early/mid 60’s mono singles, which it will mainly get used for.

Terry
Valvepower is offline  
Old 30th Mar 2017, 1:49 am   #14
Paul_RK
Dekatron
 
Paul_RK's Avatar
 
Join Date: Nov 2003
Location: Fakenham, Norfolk, UK.
Posts: 4,244
Default Re: Spot the UA15 Black Box 1004 part!

Quote:
Originally Posted by Edward Huggins View Post
The true "Black Box" cabinet was only ever used on the very first Pye Black Boxes (EL42s) from 1954 to early 1956. The Chinoiserie decoration in the greater part, and to the best of my knowledge, were Transfers. Some got distorted upon application.
Hello Edward,

A little off topic for this thread, but I'm pretty certain these weren't done by transfer. I have one here and have compared it with photos of others I've seen, and no two have even nearly the same decoration - elements of the designs are differently arranged, painted differently or just missing, and the whole image area varies considerably in size. Here's yestertech's Box copied from a 2008 forum thread on the left, mine on the right. Looking at the forms of the left hand pagodas, or the types of vegetation immediately to the left of them, or the overall size of the decorated areas... they're so different one from another that I can only think the designs were produced freehand from a very basic pattern indeed.

The black form of the transistor Box comes as news to me too.

Paul
Attached Thumbnails
Click image for larger version

Name:	twoboxes.jpg
Views:	111
Size:	40.4 KB
ID:	140075  
Paul_RK is offline  
Old 30th Mar 2017, 8:41 pm   #15
Valvepower
Octode
 
Valvepower's Avatar
 
Join Date: Aug 2008
Location: Rayleigh near Southend-On-Sea, Essex, UK.
Posts: 1,852
Default Re: Spot the UA15 Black Box 1004 part!

Hi,

I did a little more fiddling about with the value of R3 and I feel the source impedance of the Acos is different (higher?) to the TX88 and therefore R3 is loading the cartridge and restricting the output. Making R3 100K is a compromise between level and loading.

As regards the Black finish I’ve had a better look and I’m confident it was done during manufacture, but I’ve noticed on an edge where the black finish has been slightly worn away you can just make out what I feel is a brown finish underneath, so I figure the cabinet started life as a Brown one. Anyway, I feel it was done by the factory, but I stand to be corrected though .... It doesn’t look like an amateur job and it seems strange someone would buy a Black Box and disassemble it and get it finished in black.

I can only assume they were made to special order for a dealer or shop? Maybe Tenth anniversary edition of some sort?

Regards, Terry.

Last edited by Station X; 2nd May 2022 at 9:30 am. Reason: Dead link removed.
Valvepower is offline  
Old 31st Mar 2017, 8:40 am   #16
Edward Huggins
Dekatron
 
Join Date: Sep 2015
Location: Southwold, Suffolk, UK.
Posts: 8,302
Default Re: Spot the UA15 Black Box 1004 part!

I've never seen a version of the 1004 Transistor model like this. I note also the cream colour on the inside of the lid which is always black on the mid-1950s models. This might have been a special, but I can't remember a celebratory occasion during 1963 wherein Pye would have done this?
__________________
Edward.
Edward Huggins is offline  
Old 31st Mar 2017, 11:11 am   #17
Edward Huggins
Dekatron
 
Join Date: Sep 2015
Location: Southwold, Suffolk, UK.
Posts: 8,302
Default Re: Spot the UA15 Black Box 1004 part!

Quote:
Originally Posted by Valvepower View Post
I fitted a GP91-1 to the UA15..... I’d checked the level of this cartridge and found it to be giving a healthy 170mV with a 1.2cm/sec 1K test LP, and going by the data I found on the TX88 I suppose this is as near as I’m going to get to the TX88 level.
Having thought more about your low output issue, and on re-reading the thread, I notice you refer to a Collaro/Ronette TX88. If this was the cartridge that was originally fitted then this was an unusal choice as it was the cartridge used on the (much sought-after) 400TP Collaro Transcription deck. Somewhere do I have the specification for it.

Whilst the Acos GP91/1 should be ideal (and I know I endorsed it earlier) if the low output issue persisits, you could try another stereo-compatible mono cartridge, the BSR X5M, which does give around 250 mV, which is higher than the Acos.

I had a low output on my 1004 and it proved to be a PSU problem.
__________________
Edward.
Edward Huggins is offline  
Old 2nd Apr 2017, 7:50 pm   #18
Valvepower
Octode
 
Valvepower's Avatar
 
Join Date: Aug 2008
Location: Rayleigh near Southend-On-Sea, Essex, UK.
Posts: 1,852
Default Re: Spot the UA15 Black Box 1004 part!

Hi,

I’ve double checked the PYE service data on both Pauls VRSD Disc and the R+TV manual and it states Ronette TX88. Going by the info on Ben’s cartridge ‘sticky’ the Ronette TX88 has an output (ulostulo to use the Finnish) of 158mV and according to the carton the Acos GP91-1SC has an output of 200mV.

I measured the actual Acos cartridge using a 1.2cm/sec Decca LXT2695 test LP and a Levell TM3 meter and it measured 170mV so your suggestion of the GP91-1SC was absolutely fine, but for some reason it just fell short volume wise, ether because they didn’t actually use a TX88, which I rather doubt, or the source impedance was different to that of the TX88 and the input circuit R2 and 3 loaded the output of the Acos cartridge?

Your suggestion of using a medium output BSR cartridge is a cracking idea – but where was I going to find one without re-mortgaging the house!

Thankfully I’d picked up an HMV 2041 at last October’s Audio Jumble for a site less than the cost of a cartridge! Well, this uses the BSR X3M, which is the sort of cartridge I’m after – so removing the X3M from the HMV and fitting it to the Black Box – hey, presto all the volume I need plus the loudness ‘feeling’ to the volume was there.

The only problem I have now is that I have an HMV less a cartridge, so I fitted the Acos GP91-1SC from the Black Box to the HMV and it worked fine, with plenty of volume. The HMV has an orthodox cartridge loading/matching of 1.5Meg which is a better match for the Acos than the Black Box?

The other good thing looking through the cartridge bits and pieces I unearthed a Sonotone 3549 I’d forgot I had.

Terry.
Valvepower is offline  
Old 2nd Apr 2017, 8:43 pm   #19
Edward Huggins
Dekatron
 
Join Date: Sep 2015
Location: Southwold, Suffolk, UK.
Posts: 8,302
Default Re: Spot the UA15 Black Box 1004 part!

Your cartridge swap has worked out well and, yes, each cartridge is a good match for either player. Just make sure the tracking weight is optimised. The BSR X5M is now getting very low in stock (everywhere) and is priced at around £30/35.
__________________
Edward.
Edward Huggins is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 9:52 am.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.