PYE P75 radio conversion to guitar amplifier

[Gabe001]

If you need a replacement, the el41 likes a 7k load at 250v. It's an unusual impedance for modern day transformers. As you're running at 200v, a more typical 5k load will probably be OK. Alternatively, use a 3k or 3.5k and attach an 8 ohm speaker to the 4 ohm tap to double the impedance.

[Boulevardier]

Forgive my ignorance of valve theory, but can you explain how V4 can have virtually zero anode load (7.5-ish ohms), yet have a normal cathode current of about 32mA (see posts 75 and 76)? I'm confused!
Mike

[RogerLLL]

Originally Posted by Gabe001

Just for completeness could you remove/snip off c23 and re-measure?

If you want to play loud, I would suggest more powerful output valves and suitably rated transformers.

Fair comment. I really was just trying it, and now know the amp's limitations, so so long as I stay within those I should be ok - provided I can get it working again. I suspect a higher output amp would be another project.
The resistance reading across the terminals without the cap went up from 1.3 to 541 ohms. That said there seems to be some instability in the measurements. The resistance across the other terminal and the split was the same at just over 6 ohms.

[Gabe001]

6 and 541 look much better than 6 and 1.3!

Check the cap you removed with a multimeter to see whether it was indeed shorted

[RogerLLL]

Originally Posted by Gabe001

6 and 541 look much better than 6 and 1.3!

Check the cap you removed with a multimeter to see whether it was indeed shorted

The cap was toast! Now replaced and the amp's working again! Thanks for that!
The OT does meet the spec as in the attached:

[Gabe001]

Success! Well done.

[trobbins]

In an amplifier with an SE output stage, where the B+ supply has significant ripple voltage, a small portion of the output transformer primary winding can be used to neutralise (humbuck) the ripple voltage seen across the main portion of the OT primary winding. That is likely that was designed for here.

In practise only a small % of OT primary winding is required, and as an example Hammond type 52 and 62 OTs provided a 3% tapping. Neutralisation aims to equate the winding turns and ripple current products (N.I) for each portion, and the ripple voltage magnitudes seen at each end of the primary winding due to ripple current (where anode resistance by ripple current in that portion equates to RC filter resistance by ripple current into preamp portion of the supply). So for a 40kΩ plate resistance, and 1kΩ filter resistance, a 1/40 = 2.5% tapping is anticipated. The winding turns and DC current product (N.I) through that minor winding provides insignificant offset to the DC idle bias current in the SE output stage, so there is no benefit for core saturation effects.

What was the value of C23. A Hammond example used 2nF. That capacitor needs to have a suitably high AC voltage rating out to a frequency beyond the audio range.

[RogerLLL]

I've fitted a 4nF rated at 250vac, not sure what the DC rating would be. It's described as a tone corrector cap, so was intending to try a couple of different values to see what effect it might have on tone, maybe a 1n and 10n.

[Robert Gribnau]

Originally Posted by Boulevardier

Forgive my ignorance of valve theory, but can you explain how V4 can have virtually zero anode load (7.5-ish ohms), yet have a normal cathode current of about 32mA (see posts 75 and 76)? I'm confused!
Mike

The measured cathode current is a static value. As long as there is sufficient B+ dc-voltage at the screen grid (since in pentodes it's the screen grid voltage, together with the bias, that are decisive for the amount of anode current; the anode voltage is of way less importance for the amount of anode current) and a dc-path from the anode to B+, the valve will behave normally if you only look at the static conditions (so without any ac-signal).

Originally Posted by RogerLLL

I've fitted a 4nF rated at 250vac, not sure what the DC rating would be. It's described as a tone corrector cap, so was intending to try a couple of different values to see what effect it might have on tone, maybe a 1n and 10n.

The function of C23 is to prevent ringing in the output transformer. Since with signal the voltage over the primary can swing between close to 2 x B+ (so close to around 400 V, and with lower quality output transformers like in these radios sometimes spikes of even higher voltages than this 400 V can occur) and 0 V, the voltage rating of the 4nF capacitor you fitted, is way too low. It should be 630 V or even higher to be safe.

What I don't understand is why the dc-voltages got higher after C23 shorted.

[Radio Wrangler]

The tone correction capacitor is bridged across the main part of the primary winding of the output transformer. This winding has an awful lot of turns of fine wire in order to get the turns ratio to the output winding. The turns ratio transforms the impedance of your speaker to an impedance presented to your output valve which better suits the voltage and current capabilities of that valve. This ratio and your speaker impedance therefore optimise the amount of power you can get from that valve. Any other ratio, whether higher or lower, and the power available will start to drop away. It isn't an abrupt loss, but you have some tolerance before you start to notice it.

This long length of thin copper wire has resistance. At the current taken by your output valve there will be some drop of DC voltage. If the tone correction capacitor across this winding fails short-circuit, then it provides a lower resistance path for the anode current of the valve, and shorts the DC voltage drop of the winding, and you measure a higher anode voltage.

You mentioned a higher power amplifier...

Yes this is really a quite different project, but all you learn on the lower power, first toe in the water, job applies.

You'll need to move up a rung in terms of valve families to something more powerful. Also the single-ended architecture places some heavy requirements on the DC saturation capability of the output transformer. The DC quiescent current your output valve is biassed to acts to create a level of continuous magnetisation of the transformer core.

Magnetic cores aren't nice and linear. They can saturate and handle no more magnetisation, creating a limit on power and a lot of distortion. With the quiescent current of a single-ended amplifier, the DC current component biases the core closer to saturation in one direction than it is in the other direction. This has a significant effect in reducing the power level that a given size of output transformer can handle, and gives you sudden onset of asymmetrical distortion. So upping the power forces you rather rapidly into very expensive, very large output transformers. Your output transformer can need to be even larger than your mains transformer! (Mains transformers don't need to carry DC)

So, for a higher power amplifier, there is a lot of pressure to go to a push-pull architecture. You get the power capacilities of two of your chosen type of output valve, and more importantly, the DC current components of the two valves flow in primary windings wound in opposite directions... the DC magnetisation from one cancels that of the other, and the threat of asymmetric saturation goes away. You aren't pushed into disproportionately huge transformers (and their expense) and you can get the full power capability of the transformer you get.

You'r usling the EL41 at the moment? That is a small valve, used in radios made in a relatively short era. No more are being made and as the EL41 is well known as not having a particularly long life, they have been in demand for restoring radios. Stocks of remaining good ones are running low and their prices are srarting to rise.

Guitar amps give valves a hard time. So you want to choose a valve type that you should be able to get reasonably easily in the future when you need replacements. So, the EL41 with low stocks and no current manufacturer isn't a good example. That it also has a reputation for wearing out quicker than many other types confirms it as an example of what to avoid.

So, the next question has to be 'How much power do you want?' It's a difficult question. The obvious answer of 'Lots' or 'infinite' miss out on the fact that many guitarists like to drive amplifiers into limiting, giving compression, distortion as part of the performance. If you have too powerful an amplifier for the venue where you play, then you can become excessively loud before the effects you want to achieve come into play.

This is why guitar amps come in a range of power levels from practice jobs to arrays of monsters for filling a stadium.

Your modified Pye radio is right at the practice-amp level.

Looking at power levels versus valves, they come in families. Let's try collecting them together:

6V6, EL84
6L6, 807, KT66
EL34
KT88
813

The 6V6 is used in some of the lower powered Fender amps. They have a great reputation for robustness and long life. Lots of NOS ones available at reasonable prices.

EL84 are about the same sort of power, but in smaller bottles. They don't go into distortion as gently as the 6V6 and don't last as long. Currently made, but opinions differ on how good new ones are compared to period ones. Hifi people are competing for whatever is available.

6L6 is the 6V6's larger parent. A step up in power, but similar in characteristic curve shapes.

KT66 is a British valve similar to the 6L6. Reckoned to be lower distortion and used in some classic hifi amps, so the hifi crowd are paying big money for period examples. Some new ones are being made, but people are not entirely convinced of the quality.

807 is the higher power radio transmitting version of the 6L6. Ceramic insulators inside and a top-cap anode connection lets it go to somewhat higher voltages. Not being made, but lots floating around. Not quite indestructible but getting there.

EL34 Very popular in larger guitar amps, Responsible for the reputation of power and hardness of many hard rock performances. A pair or quad of these in Marshalls. Still currnetly made new, but the usual distrust floats around about quality/longevity of newly pade ones. Fairly robust but don't last forever.

KT88 A step up in power. Currently made and expensive because the market will stand it. Preferences for period ones has made their prices extremely expensive. A pair of these will do 100W easily. In some classic hefty amplifiers.

813 these are scary. 400W easily from a pair, 2000v power supplies. No longer made but plenty around. Not very expensive. Oh, and those high voltages are on anode top cap connectors. I did once build a 1000W beast with an array of these things. Transformer winding insulation was a major hurdle. Don't go here!

There are other valves, but they are mostly related to these classics.

6BW6 is a 6V6 crammed into a smaller bottle
5B/254M is an 807 also crammed into a smaller bottle.

All the different types were not developed in isolation. One manufacturer would see what an other was offering and needed to get something into that market. They were competing hard to get designers to design their parts in, knowing that users would be forced their way for replacements.

Anyway, that's enough waffle.

As a next project, I'd suggest something with a pair of 6V6 as a definite step up in power, but also good for reliability.

Everyone will have different opinions, so look for patterns where advice seems to converge.

David

[Sideband]

Originally Posted by RogerLLL

I've fitted a 4nF rated at 250vac, not sure what the DC rating would be.

If 250 is the AC (RMS) rating, multiply that by 1.414 and you get the peak (DC) rating (353 volts). Personally I would would double that for the peak - peak rating so you are looking for a cap rated at 750V DC or higher. Typically most tone correctors are rated at 500Vdc or 750Vdc. A modern equivalent at 650Vdc would probably be OK but I would err on the cautious side and fit a 1000Vdc component.

[Gabe001]

Yes a 1kv ceramic (or higher) would do. I would stick to the original value as it serves the dual purpose of shaping the high frequency response, possibly preventing ringing and instability depending on the quality of the output transformer and it's resonant frequencies.

So if the original is 4nf, the closest is 4.7nf

If you want to built a more powerful amp, as radio wrangler suggested, a 6v6 push pull amp would represent good value for money. I buy 3mm aluminium plates and drill them myself with a standard mains powered drill and an HSS stepped drill bit, and bolt this down onto a wooden surround. There is no shortage of proven designs and schematics available online for guitar amps

One final thing, I noticed that there is no coupling electrolytic cap across the 220ohm cathode resistor on the output valve. I've never seen this before and wonder whether I am missing something?

[RogerLLL]

Tbh I was thinking of a push pull of about 20w. I did used to own a ProAmp Viper, a hybrid 60w amp with a pair of 6L6 tubes as power amp & solid state pre-amp. Really was great sounding but too powerful for what I needed & RIDICULOUSLY heavy! 2x12" celestions didn't help on that score! So for small gigs mic'd up to a PA a lower wattage amp would be better, which can be driven to decent overdrive without too much volume. So really not sure what valves would work for this setup. 12AX7s for pre-amp for sure, but not sure what power tubes could work in a push pull setup? I did consider trying a single EL84 in place of the EL41 on this amp, new base obviously, which would up the power a little. The nominal voltages etc don't look vastly different from the EL41, but would this be too much for the OT?

[RogerLLL]

The Harley Benton Tube 15 amp gets amazing reviews & is about what I'm aiming for. It uses 2x EL64s so might be an interesting model to emulate:
https://harleybenton.com/product/tube15-celestion/
That said, I'd be keen on 2 channels so a variation on this theme might work.

[Gabe001]

I don't think an el84 will make a noticeable difference in output volume. I think you've done the best with what you've got, and you've done well imo.

A push pull amp using 6v6gt valves will yield you about 12w, give or take, depending on transformer losses. If you use two 6v6s from JJ you can run them at higher voltages that standard 6v6gt's and probably get close to 20w from a push pull pair. Distortion will be high-ish at max output, but that's not a problem usually for a guitarist. Beware, this is addictive.

[Robert Gribnau]

Originally Posted by RogerLLL

The Harley Benton Tube 15 amp gets amazing reviews & is about what I'm aiming for. It uses 2x EL64s so might be an interesting model to emulate:
https://harleybenton.com/product/tube15-celestion/
That said, I'd be keen on 2 channels so a variation on this theme might work.

The "two EL64s" in the description on the Harley Benton site is a mistake since the EL64 doesn't exist. Under the specifications they correctly write "Power amp tubes: 2x EL84".

[RogerLLL]

Originally Posted by Gabe001

I don't think an el84 will make a noticeable difference in output volume. I think you've done the best with what you've got, and you've done well imo.

A push pull amp using 6v6gt valves will yield you about 12w, give or take, depending on transformer losses. If you use two 6v6s from JJ you can run them at higher voltages that standard 6v6gt's and probably get close to 20w from a push pull pair. Distortion will be high-ish at max output, but that's not a problem usually for a guitarist. Beware, this is addictive.

No kidding! I'm hooked already! I'll look into the option of two 6v6s, sounds like it could work.

[Gabe001]

Originally Posted by RogerLLL

The Harley Benton Tube 15 amp gets amazing reviews & is about what I'm aiming for. It uses 2x EL64s so might be an interesting model to emulate:
https://harleybenton.com/product/tube15-celestion/
That said, I'd be keen on 2 channels so a variation on this theme might work.

EL64 valves don't exist. They mean el84, almost certainly pentode push pull. I have a 12ax7 el84 push pull stereo amp myself (hifi, not guitar - see pic) and it sounds great so no concerns from my end if you want to go down this route.

[Boulevardier]

Originally Posted by Robert Gribnau

The measured cathode current is a static value. As long as there is sufficient B+ dc-voltage at the screen grid (since in pentodes it's the screen grid voltage, together with the bias, that are decisive for the amount of anode current; the anode voltage is of way less importance for the amount of anode current) and a dc-path from the anode to B+, the valve will behave normally if you only look at the static conditions (so without any ac-signal).

Thank you for that very clear explanation, Robert. I did wonder if, as you've explained, the screen-grid voltage was implicated in some way. I'll clearly have to brush up my valve theory!

Mike

[RogerLLL]

Originally Posted by Gabe001

EL64 valves don't exist. They mean el84, almost certainly pentode push pull. I have a 12ax7 el84 push pull stereo amp myself (hifi, not guitar - see pic) and it sounds great so no concerns from my end if you want to go down this route.

Looks nice! This could be a good route to go down, I'll see if I can find a schematic for the Harley Benton amp as a start point.