Old PA - Lorenz SMV4/8W

[Herald1360]

You won't get anything out of a valve rectified PSU until the mains input is high enough to wake up the rectifier heater. 80V may simply not be enough- it'll only provide 1/3 of the nominal heater volts or only about 1/9 of the heating power (give or take- the cold heater resistance will be a bit lower than when they're up to temperature).

[trobbins]

Does the PT only have one primary winding with no taps?

If no taps, then I'd recommend removing the rectifier valve so that no loading is on any secondary, and use the variac to bring up the primary to your nominal mains voltage, and measure the heater voltage, and a half-secondary HT voltage. You need to be confident in your meter and probes, and preferably have everything in place before energising, and only move probes when un-energised. Do you have earth fault cb feed (or similar), and fused chassis, and protective earth fitted to chassis? The heater is likely to be 10% high for no load and nominal mains, but I've seen higher. It's best to know what the unloaded voltages are for starters.

To test under loaded conditions, then I'd recommend making up a dummy solid-state rectifier valve base (2x 1N4007 in series to replace each valve diode), or temporarily soldering 2x 1N4007 in series from pins 4 to 8 and 6 to 8 on the rectifier socket. Then use the variac along with what ever loading resistors you may have. That can test the B+ levels around your circuitry, and the coupling cap leakages, but you have to be careful not to exceed any coupling cap or filter cap voltage rating. You would have to be well kitted out with a variety of load resistors to test valve amp B+ supplies - I have a few suitable power resistors that I interconnect with spade lugs in a careful loop on the bench just for that purpose.

There are some testing notes in link:
https://www.dalmura.com.au/static/Re...0PA%20amps.pdf

[Richardgr]

Hi Tim, that treasure trove of useful information is already one of my bookmarks, and now I see that you are the author! Thanks for the link.

The transformer has no markings, but by checking resistances I had found another tap for a higer input voltage. When I checked that with only the PT and Variac, I was only able to get to 3.8 for the rectifier heater (should be 4V), and 6.1 for the other tubes, and 562V across the HT.

The lower input voltage tap gave 4.2V, 6.8V and 624V respectively, so looked to be a good bet unloaded.

I'm fairly sure that I could get the amp operational again, but it is a bit indulgent to have two EF12's and an EL11 for someone to play their guitar, and I had higher ambitions for the chassis.

Probably like many who post here, I've acquired too many projects, and I really need to clear the decks a bit. I'm concerned that without a good plan for this amp, I'll end up devoting too much time on it.

I'm going to have one more shot at testing, with the other tap this time. As advised, I'll substitute silicon diodes for the rectifier. Then I think I need to put the project on ice.

I have been reading the Linear Audio articles by Frank Blöhbaum on MTA (Multiple Transconductance Amplifiers), and I have seen some people have used the Y8 based ECL11 for this application, so I think that is going to be my aim for this chassis. I just need a couple of small SE OPTs, and the tube count can be 2 or 3, depending on rectification. I have acquired a bunch of UCL11's, so am thinking a transformer with a 120v secondary will provide serial heater supplies and B+ (with a doubler), and their should be room for a toroid under the chassis.

The transformer that is fitted is not original, and the OPT is small and is fitted under the chassis. I haven't drilled or cut anything, and all parts are saved and marked, so my options are still open.

[Richardgr]

Originally Posted by Herald1360

You won't get anything out of a valve rectified PSU until the mains input is high enough to wake up the rectifier heater. .

I had measured volts unloaded with 80v supply, and then when loaded, rectified volts across the load were not increasing, so I just lost confidence in my approach. I plan to rerun the test with silicon diodes.

[trobbins]

The main secondary is likely a HV-0-HV format. Only measure the 0-HV voltage, otherwise you may be measuring twice the HV voltage.

[Richardgr]

When I measured before, it was 281-0-281 with the 245v input tap, and 312-0-312 with the 220v one. Our electric supply is exactly midway between those taps, so it is a bit awkward with the heater supplies. The rectifier requires 4V, and it would be 3.8V (unloaded) with the 245V tap.

I have made it difficult for myself since I only have 16uF capacitors with 350v rating. I will order some more caps so I can test without worrying about exceeding ratings.

[trobbins]

Perhaps try and make a measurement of heater voltage with the AZ11 heater only in the 4V circuit, and the other valve heaters in circuit for the 6V3 circuit, to confirm the actual voltage at the socket terminals for the 2 primary taps.

Do you know if your mains voltage varies over a day, or over time?

In practise it may be easier to drop the heater voltages a titch to get them right and live with the generated B+. The only other proviso is that primary magnetising current does not increase a lot when on the 220V tap compared to the 245V tap.

[Richardgr]

This small little amp, that looked so simple, has turned out to soak up a lot of my time, as I try to make sense of the schema.

I had a bit of a shock when I took a census of the projects I have on my plate, and I decided that I need to move some things on to a state where I can pass them on. This amp was a death trap of crumbling wires when I bought it, but the basic setup is not a million miles away from a Fender Champ style amplifier, so I decided that I would aim for a circuit as close as possible to the original, reusing the resistors on the paxoline board, since they all test OK, and package it as a guitar or microphone amp.

I am dying to hear how it sounds! The power supply has a nice choke, and the singled ended OPT is not as small as some Champ ones, so should be an interesting combination, with that EL11.

However, I can get about 90% of the schema, but I just cannot get it into a state that makes sense, that I consider 'definitive'.

Before I dismantled it, I captured all the wiring connections in a DIYLC wiring diagram. I double checked the connections too and from the components on the paxoline, and then I committed to taking the decaying capacitors away, and cleaning up the wiring and sockets.

I have emptied the 16uF + 16uF / 500VDC can caps, and restuffed them with long life, high temp variants. That was very satisfying - it looks nice and original, and keeps extra real esate under the chassis.

The wiring diagram and my latest schematic are attached below. It is obviously not OK because R15, 700K, goes from an earth to another earth. There is some fiendish sort of feedback at play, because there is no disputing that 50nF cap that goes from the anode of the output tube back to the driver tube output.

I have struggled with interpreting and drawing the tone control circuit, which consists of a Bass cutoff switch, and a tone control. Maybe the output tube anode cap is linked to that in some way.

If anyone can give me some pointers, that would be great, otherwise I will try to wing it and build what I understand, using a simple tone control layout, and leaving the feedback out of the equation.

The circuit included is from a Russian website, presumably from an old document, which shows a similar sort of feedback setup, for information.

Richard

[trobbins]

Remove the link across R15. The frequency response of the various sections of circuitry (including feedback) can be deduced by frequency spectrum sweeps, or calculation, but aren't going to stop the amp from working. Use a resistor load instead of speaker (unless it is a beefy speaker) in case of instability. Working through any undocumented circuitry takes time and measurement and instrumentation and a dollop of theory. You can always short across parts (eg. R10), or open connections (eg. feedback), to simplify operation when testing.

[Richardgr]

Some progress.

I have 400v from the rectifier unloaded. Modelling in PSUD2 I get a theoretical B+ of 258v after the choke, which has a resistance of 245R. (I have used 15H to modelling the choke - is that reasonable?)
With cathode bias the EL11 should just be in spec.

Aiming for driver and output stage, testing with Bluetooth receiver.

[trobbins]

The choke resistance of 245R indicates is is likely to have at least a 100mA rating (2.5W), and perhaps up to 150mA (5.5W). For that physical size it could be in the range 10-25H I guess. One way to test is in link below.
https://www.dalmura.com.au/static/Ch...easurement.pdf

[Richardgr]

At last some progress. I felt unable to build anything unless I had a schematic, and I was reasonably confident it would work. Firstly, I started to try to finally get to grips with LT Spice. I had managed to simulate a preamplifier I have, and I was amazed how closely the theoretical values correlated with the measured.

I tried to derive a working simulation from the pieces of the schematic that I had derived, but I was stuck with models for the EL11 and EF12. I managed to get a model for the EF12 from DiyAudio, and the EL11 has very similar curves to the EL41, which I had.

The original schematic had one of the EF12's triode wired, and I was not successful modelling that, and spent a lot of time trying every permutation of components from the schematic I had, without getting anything to work.

Part of the problem is emulating the output transformer, and I found a good model for that of another site. Even with that, I was unable to get anything to simulate.

Then a stroke of good fortune. Searching for 'EF12' and 'amplifier' in Swedish, I came across this site:
Kungs Radio

The page linked is a list of custom schematics from the 40's, using the same tubes. I could see the design patterns, and I was able to find a very simple schematic which would be a good basis for what I was trying to create - a Champ style amplifier.

I tried to simulate that, and once I realised that 1m for the resistors meant 1milli-ohm, and I had to write 1Meg, then I finally got a nice sine wave output at the speakers.

(Any comments on the schema greatly appreciated).

The picture is work-in-progress.

[Richardgr]

I feel I have come to the end of the road with this project, so this is the final status. Music was achieved, albeit far from hi-fi.

The circuit is in the PDF below. I decided to make the first EF12 triode connected as a means to reduce gain, since I had gain in spades. The circuit I was building had feedback to the driver tube, so I would lose some gain there, but it all looked OK on the simulator.

I have learnt a lot trying to make some headway with this thing. Firstly I was too quick to dismantle it. I should have drawn and understood the schematic before dismantling - instead I used DIYLC to capture the connectivity, and I made a mistake somewhere with that, which I could never unravel, so that left me with a set of parts but no firm plan how to build them. Secondly, it forced me to get to grips with LTSpice, as a mechanism to evaluate a design and create a baseline. EF12 Triode models were unavailable, so I learnt how to import the curves from a data sheet and make my own tube models, which was quite straightforward.

Now I have a much greater respect for the designs of the 40's. There is a lot of stuff going on in the original circuit - subtle feedback and tone arrangements - that one does not see anymore in modern schematics. The components probably dictated some of the design ideas, but one gets the feeling that there was a 'feel' for design that has been replaced with a simulated, empirical approach today.

I had some issues that betray my inexperience I still have with this hobby, so if anyone can help with these silly questions, it would be much appreciated!

Firstly, I could not get the EF12 to ever exceed 10v on G2. I thought if it was triode wired (screen tied to anode) that would make it easier to unravel, but then I could not exceed 10V on the anode and screen. I do not have a base for testing the German steel (Y8 base) tubes, so that was an uncertainty: I never knew if the tubes wer any good in the first place. I will plan to create a solution so I can test all the tubes.

The circuit I was copying had a 1Meg log volume control and tone control, but the original amplifier had 100K, so I substituted those. I think that was one reason for the poor performance - the first EF12 was not able to drive the 100K properly. I noticed the example schematic in the data sheet had 700K after a coupling capacitor.

I was very happy I had splashed out on an oscilloscope, and I was able to get more experience with it. That helped me identify oscillation due to the feedback. I just could not get the feedback to work, regardless of which way the OPT was connected, so the simulation did not help with that. (Oscillation was around 10-12khz from the 2nd EF12).

When I had oscillation issues, the outpt tube grid voltage measured -50VDC. Why is that?

Also, I had something very strange. Depending on which end of the wire I tested, that connects the anode of the output tube to the OPT, I got a different voltage! 258VDC at the OPT - which was reasonable - or 296VDC on the valve pin - which was not. What basic thing am I missing there?

Anyway, now it is back together, and it not really any closer to being useable. Time for me to reappraise what to do with it.

[trobbins]

If you get keen again ....:
Stages of the amp can be tested in isolation. Eg. input a low signal level to the input and 10:1 scope probe across VR2A. Disconnect C10 and input a signal across VR2A and 10:1 scope probe across R11 (with EL41 removed). Disconnect C6 and input a signal across R11 and scope probe across speaker resistor load. That is one way to confirm if each valve is functioning assuming anode DCV of U1 and U2 and screen voltage of U2 is ok.

You should also look at how wires are located under the chassis. Eg. speaker wire should be twisted and run together from OT terminals to speaker terminals; EL41 anode wiring should be kept as far from any other wiring as possible (ie. lay it along the chassis and keep other wiring distant, and connect C12 at the socket); keep HT and first filter cap leads direct and twisted and as far from circuitry as possible; keep tone and vol pot wiring shielded (as this is likely where your oscillation is being caused from stray capacitive coupling in to high impedance nodes).

[Richardgr]

Hi Tim,

Thanks for your support with this, and the good tips you have provided. Once I have tested the tubes I will revisit it, but I need to make up some socket adaptors for testing, and I may have to work abroad for a few months in the near future, which has upset my plans.

What would be your goal with this type of project? Looking at your site, do you find homes for the PA amps that you restore? I had thought if the EF12's test good, then a mic amplifier might be an option. Alternatively, using the chassis with a couple of ECL11's as a headphone amplifier could be an idea.

I am annoyed with myself that I lost the option of restoring the old circuit. There must be a schematic somewhere out there

Rgds,

Richard

[trobbins]

Vintage equipment restoration can sometimes take many years to come across original schematics or someone with same equipment.

I generally keep mine 'on the shelf' for future reference, as for me this is about the technology rather than a final target use.