Leak TL10 power supply modification

[qualityten]

I have a TL10 which originally came from a Bal Ami E80 jukebox.

I received it in very dirty condition, with some repairs to the tag board. More significantly, it was originally a US 117V version. At some point, the power transformer was replaced with a 240V one with diodes in place of the valve rectifier and no centre taps for the secondary HT and LT windings.

As the HT was low, at 250V in each side, I decided to replace it with a transformer from a guitar amp which runs two 6L6s. However, even this only supplies 300V on each side.

As I want to use it as a standard TL10, I have removed the volume and tone controls and the relay.

Now my question: As I do not have any KT61s, ought I to be able to get it running with 6L6s on a slightly lower voltage supply? If so, can I just lower the value of the 1K 10W resistor between the first two filter caps? I have replaced the four 8uF caps with a four section 20 20 20 40 can and am thinking of using a GZ34 in place of the original 5V4 for less voltage drop. I could also consider going back to silicon diodes for even less drop.

And, as I don’t have another to compare, please can someone who owns a TL10 tell me where to locate this resistor? I can see it as the large green one evident in pictures.

I’d be grateful for pointers from those who know these amps well. Thanks.

[snowman_al]

I have not held a TL/10 for a few years now and never seen one that was used in a juke-box. Could you post pictures?

You can 'officially' use 6L6s in there if you change the cathode bypass resistors (R14, R24) from 220 ohms to 600 ohms. Also replace R3 and R5 on the EF86 for new values. All this is on the circiut diagram see here: http://44bx.com/leak/Leak/Circuits/TL10both2_cct.gif

The large 1k resistor (R20) from memory mounts (originally) on the 5 volt tap on the mains transformer to the tag strip. If your transformer is not configured the same, you will need to add a tag strip for support.

I'm guessing the 300 volts you refer to are AC? That will get you 420 to 430 volts DC off load. How much that drops will depend on the efficiency of the rectifier, as you say, and the current rating of the replacement transformer.
You will need to get it working then adjust the value of R20 to get about 330 volts on the 6L6 anodes.

[qualityten]

Thanks snowman_al. I've noted the small changes needed to run 6L6s, but was wondering if adjusting the value of R20 would affect the stability of the power supply.

I will post 'before and now' pics. The original power transformer shed all its wax onto the wiring loom, so I had to untie parts of the lacing and clean up a lot before I could make out wire colours and resistor colour bands.

By the way, I admired your restoration of a Beam Echo/Avantic integrated amplifier in PFM a few years back. Beautiful work!

[Herald1360]

The ratings for the cathode resistors in that reference don't add up.... why would you need individual cathode resistors of 600R 5W each or a common one of 300R 3W? Either the individual ones are well overrated or the common one ditto underrated!

[qualityten]

I agree. Does the omission of the bypass capacitor make any difference?

[Radio Wrangler]

Without an original mains transformer, the value is going to be rather impaired should you ever decide to sell it, so you might as well do what you want with it.

Re 6L6s: These are available, cheap and have a fine reputation for robustness. Your HT might not be ideal and the output transformer may be a bit off in ratio, but unless you want the last bit of power out of it, these are all survivable and as Snowman says, you need a resistor change to suit the difference in grid characteristics.

With just one amplifier, you don't need to create a matched pair.

David

[qualityten]

Thanks David. I do not have a lot invested in it and so am not concerned about value, though I intend to restore it sympathetically. The chassis itself has a few extra holes associated with the original tone controls and volume pot for jukebox use. The extra EF86 added for tone control was mounted on a plate under the octal socket, so it's been possible to drop in an octal socket without disturbing the chassis further.

The output transformer is original, and if a replacement power transformer were to come on the market I'd consider restoring it further by adding the original.

[qualityten]

I attach two 'as received' and two 'as now' pictures. The internal component replacements are just what I have to hand. I intend replacing these with more suitable ones, e.g. 2W resistors. The dropper resistor also needs to be moved to the side of the chassis.

[qualityten]

A quick update, if I may. I've got the amp working. A 150R dropper resistor supplies the correct voltages for 6L6 anodes. And 220K (not 330K) for R5 gets the right anode, screen and cathode voltages for the EF86

But I can't get the bias voltage right for the 6SN7 and the 6L6s For the 6SN7, the anode voltages are 25V too low and the cathode voltage is 9.3V instead of 49V. For the 6L6s, while the anode voltages are okay at 322V, the cathodes are 26V instead of 75V. These are going to one resistor, without a cap.

Can anyone advise on how to get the cathode voltages up? I presume in each case the lower voltages are affecting the bias of the valves.

[Radio Wrangler]

Do you know if Leak made that amplifier with the mods for Bal-ami, or whether they Bal-Ami bought standard amplifiers and modded them?

You might have an example of a pretty rare Leak product so it might be worth stashing the bits in case you ever want to put it back to original.

David

[qualityten]

Bits have been duly stashed away. The relay has a part number inked on and varnished over.

[Silicon]

If I was modifying a valve amplifier like this, my main objective would be to run the output valves within the limits of their anode dissipation.

If you know the anode voltage, you can select a safe cathode current by choosing a cathode resistor that will give a suitable grid bias.

The schematic I found with the KT61 valves has a grid bias of 7.5V (not 75V).

7.5V divided by 220 Ohms (the cathode resistor) suggests a cathode current of 34mA for each valve.

You may need to experiment with the cathode resistors of the 6L6 valves. Start with a large resistor and adjust it downwards to increase the cathode current.

[snowman_al]

Originally Posted by qualityten

... But I can't get the bias voltage right for the 6SN7 and the 6L6s For the 6SN7, the anode voltages are 25V too low and the cathode voltage is 9.3V instead of 49V. For the 6L6s, while the anode voltages are okay at 322V, the cathodes are 26V instead of 75V. These are going to one resistor, without a cap.

Can anyone advise on how to get the cathode voltages up? I presume in each case the lower voltages are affecting the bias of the valves.

I think the 6L6 cathode volts are about what you should expect. (The KT61 cathode volts are 7.5 volts not 75 volts. That gives about 38 mA through each valve with a 100 ohm shared resistor.) 26 volts over 300 ohms for the 6L6 is 44mA per valve so an easy ride for them. And Leak specify a 50 volt working bypass cap for the 6L6 dual resistor set up so are expecting a reasonable voltage.

The 6SN7 numbers are off though? What voltage do you measure across R4 and R13 in the supply chain?

[qualityten]

Silicon and Snowman, thank you for your comments on the 6L6 cathode voltage in the schematic for KT61s (7.5V not 75V--a relief)! And thank you for telling me how to work towards the right cathode resistor value.

Snowman, re the 6SN7, I'll report back on the voltages.

[Radio Wrangler]

You look at a full datasheet for the valve you intend to use. You decide on an amount of bias current and look at the graph of Ia versus Vgk and see what grid voltage will give your chosen current. Ohm's law does the rest for you. divide the voltage by your chosen current (in amps) and you get the cathode resistor value per valve. If you have two valves sharing one resistor, halve the resistor value.

This assomes the grid is biaased to ground potential and that the bottom end of the cathode resistor is ground, which is most often true.

David

[snowman_al]

Fresher eyes this morning and looking again at the pictures.

If the 'after' picture (leak4) above is how it is now, I see one error, maybe 2.
The white coffin block resistor (R9) looks to be 1K ohms? It should be 10k ohms.
Then I cannot see how the resistor (R19, the real 1k resistor) is connected? I should meet R8, R10 and R19 at the tag. The other end of it comes from the the joined pins 3 and 6 of the 6SN7. (I notice in the 'before' picture it has another series resistor added )

Alan

PS. if those Mullard EL34s came with it you have done well...

[Valvepower]

Hello,

Wow seeing this TL10 may have answered a question I’ve had about an odd-ball TL12.1 I came by a few years ago as it has similar “hand drilled” holes in the front (sadly there is no Traffolyte panel) plus it’s a 110/220V unit with a different mains transformer (is an original Leak part) to that of the standard UK/US versions.

I’ll take some pictures and post them later today.

Ps. If it helps, I have a pair of “stock” TL10’s, which I’ve restored.

Regards
Terry

[Valvepower]

Hello

As I said here are the photos of the rouge TL12.1….

As you’ll see there are hand drilled holes and change in colour to the chassis where the Traffolyte panel was sited plus an extra hand drilled hole to the side of the octal base which is normally connected to a preamplifier. I feel this carried a lead which would have connected to the top cap of what I suspect was an EF36/7 in the preamplifier octal socket.

When I got this amplifier all the extra bits had been stripped off and was in a really sorry state with an O/CCT choke, which had spewed its tar contents onto the chassis along with a completely faulty capacitor block all indicating hard use.

I feel the EF36/7 in the preamplifier socket (?) was an extra preamplifier and the holes were for some sort of controls and sockets?

The mains selector is 110/220V and a non-standard AC outlet socket which resembles a US two flat pin socket.

Now, was this an even earlier juke box amplifier?

Regards
Terry

[qualityten]

Terry, mine had an extra EF86 where your EF36/7 once was, so this may be an earlier version. But yours does look hand-drilled, whereas the holes and Traffolyte panels on mine look factory-produced.

Updating work on mine: thanks Alan for the correction to R9! With 10K there in place of the 1K, the anode voltages of the 6SN7 returned to very close to those on the circuit. R19 is present. The resistor lead links pins 6 and 3 at one end before connecting with R8, R9 and R10 at the other.

I decided to increase R20 from 150R to 280R (using an old resistor used to power to relay) and the supply voltage after R20 is now 324V. After R13, it's a bit high at 275V, and after R4 it's also high at 240V, but I think that this is caused by the increases to R3 and R5 specified by the change from KT61s to 6L6s. Am I correct?

I have yet to look at a square wave through a scope, but it sounds right, apart from a very soft popping sound only audible with an ear to the speaker.

Two questions remain:
1. Should I change the single cathode resistor for the 6L6s to one for each valve?
2. Can I check how the speaker should be connected? The schematic says speaker ground to chassis ground, but there was a piece of tape on the chassis when I got it saying that the speaker ground should be connected to the 16ohm tap. (pic attached). I initially thought this might be because of a fault with the output transformer, but it's working normally.

As ever, thanks to those who have offered advice.

[snowman_al]

Originally Posted by qualityten

...With 10K there in place of the 1K, the anode voltages of the 6SN7 returned to very close to those on the circuit. R19 is present. The resistor lead links pins 6 and 3 at one end before connecting with R8, R9 and R10 at the other.

I decided to increase R20 from 150R to 280R (using an old resistor used to power to relay) and the supply voltage after R20 is now 324V. After R13, it's a bit high at 275V, and after R4 it's also high at 240V, but I think that this is caused by the increases to R3 and R5 specified by the change from KT61s to 6L6s. Am I correct?

Good, glad it is near normal!
Do not worry about the voltages being higher than on the service sheet. They are quoted using a meter with a resistance of 1000 ohms per volt. (Today's digital meters are in meg ohms per volt range.) When measuring round the EF86 their meter would be drawing more current from the supply than the valve. And yes you have reduced that slightly too with the revised values.

Originally Posted by qualityten

... I have yet to look at a square wave through a scope, but it sounds right, apart from a very soft popping sound only audible with an ear to the speaker.

Two questions remain:
1. Should I change the single cathode resistor for the 6L6s to one for each valve?
2. Can I check how the speaker should be connected? The schematic says speaker ground to chassis ground, but there was a piece of tape on the chassis when I got it saying that the speaker ground should be connected to the 16ohm tap. (pic attached). I initially thought this might be because of a fault with the output transformer, but it's working normally.

The popping could be anything, a valve, a resistor, poor connection or... Try removing one valve at a time (starting with the EF86) and listening again. If it is still there good chance the first stage is OK, move to the 6SN7 and so on. If removing the EF86 stops the noise - fault find that stage.

1. Try both. If the output valves are well matched one resistor is fine. If they are a random mix then 2 resistors and caps are better. You will get different opinions here as to which is the better I'm sure.

2. I would use the 8 ohm tap (red spotted screws) assuming your speakers are 8 ohms. The standard TL/10 has the left hand 8 ohm connection bonded to the chassis earth. You need to look carefully at the rear of the connector, in case it is different to the standard TL/10 though. Another picture/s?