London Power ad

[-]
Search the Forum








(Advanced Search)

London Power's "PA"-series Tube Power Amplifier Kits
#1
Have you tried one of these? Post your questions or comments!


These octal, push-pull power amp kits use London Power's full-audio-bandwidth OT-10, OT-25 or OT-50 toroidal output transformers. Performance is so good they can be used for hi-fi. The power amp can be used with any of the common octal-based output tubes that use the standard pin-out. Tubes can be mixed as well!

Switching between fixed-bias and cathode-bias can be "global" (both tubes) or independent.

Lots more info at:
10W Power Amp Kit (PA66-10)
25W Power Amp Kit (PA66-25)
50W Power Amp (PA66-50)


[Image: pa66-10-kit_800.jpg]
Reply
#2
(10-27-2018, 06:34 PM)I have a question about the PA66 power output stage.The resistors R18,R19 and R26, R27  150k 1W FP- are these required for proper operation or are they optional ?I am building the circuit using a standard output transformer not the toroidal type included with the kits. Wrote: Have you tried one of these? Post your questions or comments!


These octal, push-pull power amp kits use London Power's full-audio-bandwidth OT-10, OT-25 or OT-50 toroidal output transformers. Performance is so good they can be used for hi-fi. The power amp can be used with any of the common octal-based output tubes that use the standard pin-out. Tubes can be mixed as well!

Switching between fixed-bias and cathode-bias can be "global" (both tubes) or independent.

Lots more info at:
10W Power Amp Kit (PA66-10)
25W Power Amp Kit (PA66-25)
50W Power Amp (PA66-50)


[Image: pa66-10-kit_800.jpg]
Reply
#3
Hi ironman28

Those resistors protect the OT, as TUT describes. It is a method that as far as I can tell is one used only by me and readers of my books.

have fun
Reply
#4
Which volume and chapter is this in? I guess I must have missed it and would like to read up on it. Thanks!
Reply
#5
Hi Maikonrose

That's a good question Big Grin

Frankly, I cannot find it, although I did discuss it on the PowerScaling forum back when that was active. Gingertube then took it up in his amps and reported about it on other forums.

Basically, the resistors to ground from each end of the OT are a benign form of energy limiting and absorption in the case of a discontinuous load. The resistors also help stabilise the amp against oscillation by working against parasitic capacitance and inductance.

The modern use of reverse-wired diodes across the tubes is a bit brutal. Two 1N4007s provide 2kV of reverse voltage before the diodes themselves are damaged, but it is the diodes on the opposite side that perform the protection when they are forward biased and clamp the negative peak to about a volt or so below ground, in turn clamping the reversed-biased side to a volt or so above 2 x Va. This seems a little improbable because the conducting tube cannot pull its plate any closer to ground (cathode) than the limits of Vsat dictate even at zero current.

Of course, in the above scenario if the tubes stop conducting because the reflected load has risen to infinity (theoretical but impossible), then the transformer voltage can rise similarly. The diodes shunt the energy to ground.

Using the resistors instead of diodes provides a continuous path for control of flyback energy, with no discontinuous or abrupt conduction events, so no impact on performance of the amp.

It goes without saying that any tube amp designed without a "minimum load" tied to its secondary is poorly designed. So I won't say it Smile All the TUT 3&5 project amps have such a resistor in place. This is much better in my view than having a dead short across the secondary ala Fender and their clones for when the load is disconnected. In Fender's case, the disconnection has to be at the amp end of the speaker cable for the scheme to offer any facade of protection - but a short reflects as a short and the tubes are strained even if the OT is protected. It's a choice.
Reply
#6
(07-06-2021, 04:19 PM)ironman28 Wrote:
(10-27-2018, 06:34 PM)I have a question about the PA66 power output stage.The resistors R18,R19 and R26, R27  150k 1W FP- are these required for proper operation or are they optional ?I am building the circuit using a standard output transformer not the toroidal type included with the kits. Wrote: Have you tried one of these? Post your questions or comments!


These octal, push-pull power amp kits use London Power's full-audio-bandwidth OT-10, OT-25 or OT-50 toroidal output transformers. Performance is so good they can be used for hi-fi. The power amp can be used with any of the common octal-based output tubes that use the standard pin-out. Tubes can be mixed as well!

Switching between fixed-bias and cathode-bias can be "global" (both tubes) or independent.

Lots more info at:
10W Power Amp Kit (PA66-10)
25W Power Amp Kit (PA66-25)
50W Power Amp (PA66-50)


[Image: pa66-10-kit_800.jpg]

(07-07-2021, 01:41 PM)K O\Connor Wrote: Hi Maikonrose

That's a good question Big Grin

Frankly, I cannot find it, although I did discuss it on the PowerScaling forum back when that was active. Gingertube then took it up in his amps and reported about it on other forums.

Basically, the resistors to ground from each end of the OT are a benign form of energy limiting and absorption in the case of a discontinuous load. The resistors also help stabilise the amp against oscillation by working against parasitic capacitance and inductance.

The modern use of reverse-wired diodes across the tubes is a bit brutal. Two 1N4007s provide 2kV of reverse voltage before the diodes themselves are damaged, but it is the diodes on the opposite side that perform the protection when they are forward biased and clamp the negative peak to about a volt or so below ground, in turn clamping the reversed-biased side to a volt or so above 2 x Va. This seems a little improbable because the conducting tube cannot pull its plate any closer to ground (cathode) than the limits of Vsat dictate even at zero current.

Of course, in the above scenario if the tubes stop conducting because the reflected load has risen to infinity (theoretical but impossible), then the transformer voltage can rise similarly. The diodes shunt the energy to ground.

Using the resistors instead of diodes provides a continuous path for control of flyback energy, with no discontinuous or abrupt conduction events, so no impact on performance of the amp.

It goes without saying that any tube amp designed without a "minimum load" tied to its secondary is poorly designed. So I won't say it Smile All the TUT 3&5 project amps have such a resistor in place. This is much better in my view than having a dead short across the secondary ala Fender and their clones for when the load is disconnected. In Fender's case, the disconnection has to be at the amp end of the speaker cable for the scheme to offer any facade of protection - but a short reflects as a short and the tubes are strained even if the OT is protected. It's a choice.
I don't have one these kits so I also don't have a schematic but from what I can see the 150K flameproof resistors are tied in series from the plates (pin 3) to pin 1 which connects to the bias monitoring resistor and then to  the ground? 

In speaking to a Mesa Boogie Tech one time trying to determine the type of diodes on wired as you described (one set had blown) he mentioned that at high volumes the diode had some effect on the tone of the amp.  Do you find the resistors have any effect on the tone?  

On the secondary the 270 ohm 5W like in TUT it sure seems like just a few cents save a lot of potential problems.
Reply
#7
Hi Guys

If the protection diodes are effecting the sound there is a serious flaw in the circuit - but that is not unusual for mesa products.

Every component has parasitic elements, and for diodes it is their capacitance and leakage current. A decent diode has no leakage current until very near its breakdown, and in any case, this should be microamps, nothing significant to the tube output stage. The Mesa tech is either mistaken or ignorant of the facts if he thinks the diodes change the sound.

Resistors in the same position have no impact on the sound. They have no erratic behaviour nor do they do anything abrupt. Frankly, at full output, the sound will be so loud through common MI drivers that aural compression will keep you from hearing that sound accurately, meaning that IF the diodes were to alter the sound there is no way you could tell. This is hypothetical because in the real world the diodes do no such thing to the sound.

There is nothing special about the diodes used here, either. Ideally they should be fast, so go for the UF version rather than 1N, if possible. Current handling does not have to be high but you'll see both 1A and 3A types used. Leave the leads long if you want the diodes to survive many incidents (if there are so many incidents that the diodes conduct something must change about how you use the amp).
Reply
#8
(07-07-2021, 04:33 PM)K O\Connor Wrote: Hi Guys

If the protection diodes are effecting the sound there is a serious flaw in the circuit - but that is not unusual for mesa products.

Every component has parasitic elements, and for diodes it is their capacitance and leakage current. A decent diode has no leakage current until very near its breakdown, and in any case, this should be microamps, nothing significant to the tube output stage. The Mesa tech is either mistaken or ignorant of the facts if he thinks the diodes change the sound.

Resistors in the same position have no impact on the sound. They have no erratic behaviour nor do they do anything abrupt. Frankly, at full output, the sound will be so loud through common MI drivers that aural compression will keep you from hearing that sound accurately, meaning that IF the diodes were to alter the sound there is no way you could tell. This is hypothetical because in the real world the diodes do no such thing to the sound.

There is nothing special about the diodes used here, either. Ideally they should be fast, so go for the UF version rather than 1N, if possible. Current handling does not have to be high but you'll see both 1A and 3A types used. Leave the leads long if you want the diodes to survive many incidents (if there are so many incidents that the diodes conduct something must change about how you use the amp).
I did find the notion a little strange when he told me that....   One my reactions was the same as yours: with 100W amp and EV12L how would you even hear it if it only happened with the amp cranked?

  It was a MKII which as you probably know is an utter nightmare to work on. It was designed in such a way that accessing the solder joints was hard and components are difficult to remove since they bent the lead over and left them pretty long. Even getting the chassis out the tiny cab was a hassle.  I've joked with other people that Mesa has few people on staff whose job it is to make their amps hard service.

At any rate I'm definitely going to add those resistors to new builds! Thanks for all the insight.
Reply
#9
Hi Guys

Most amp repair techs and modifiers are eager to take in Mesa amps to see what all the fuss is about - if you believed their ads these must be the best amps in the world, right?

Then you see the total absence of grid-stops, the spaghetti-like signal path, the PCB that has wiring attached from all four sides, the stacked LDRs, and all the other wonderful things that make the assembly quick and cheap but forecast nightmares for servicing or for mods. Peavey does the same things. Eventually all those techs drop mesa and Peavey before they drop any other brand. These things ARE reported throughout TUTs.

In the modern age of computerised PCB layout and overall design, it is very easy to eliminate all the wiring or very nearly so, but this means the whole amp or product is on one board. The chassis must have one panel that can be removed if one is to install or remove the card. It all looks good on the monitor and the assembly is quick, plus the board is complete so it can be tested prior to mounting in the chassis. Even though I am aware of the trap and know better, I have done this myself.

Wiring is the repair person's salvation provided it is done carefully.PCBs on the front panel, PCBs on the rear panel, maybe other PCBs for the PSU etc, wired to each other with generous lead lengths - service loops as they are historically known - allow smaller portions of circuitry to be dealt with more easily. It may end up that to replace a panel pot, say, one has to free all the front panel pots, but if this means you do not also have to free up the rear panel items or other things, then it is one step closer to being a breeze to service. Breaking the control and jack boards into yet smaller pieces makes it quicker still.

The other thing having some wiring does for you is to improve the reliability with regard to mechanical strain due to flexure of the chassis. Typically the pots, jacks and switches secure the PCB to the panel and there can be strain on the solder connections depending on how much overall support there is for the card. A further benefit to having some wiring and smaller PCBs is that each is mechanically only related to a small part of the chassis and therefore measurement and placement is less critical and tolerances of the chassis bends and dimensions is less critical. The reality is that everything is designed at once in the same software, so mechanical alignment will be as good as the chassis fabrication allows.

In the case of the PA66, the PCB is supported by metal standoffs, four around each power tube and three on the front-end. The tube socket leads are bent over for mechanical integrity prior to soldering. This is actually quite robust provided one takes the normal amount of care during tube insertion and extraction.
Reply
#10
Hi Guys

Note that PA66 can be operated with one or two power tubes, and can be expanded using PX66 to have yet more power tube positions. The tubes can be any type of the common pinout, as sets or as mixed types, and each biased or wired triode / pentode / ultralinear how the user wishes.

Remember there is absolutely no need for matched tubes, but the player can do as what pleases them to do.

The front end is optimised for 12AT7, which will provide the cleanest output. However, you can insert any 12A_7 for other tonalities and gains. A 12AX7 assists in giving a British tone.
Reply
#11
Question about the PA Series power amps: do these power amps require dc servos for the biasing to keep a balance for the toroidal output transformer like the examples in POP?
Reply
#12
Hi Guys

When using toroidal OTs, as our amps do and PA66 kits do, there is absolutely no reason to use active bias controls (servos, differential balancing, etc). Standard preferred** bias-set networks are stable over the long term.

That advice goes for EI output transformers, as well.

You do not need matched tubes either. You simply have to have a separate bias control for the push- and pull-sides of the output stage. As mentioned frequently on this forum and in our books, DC idle current balance is slightly off from the hum-balance of the OT - just a few milliamps. So, set one side of the circuit by meter for a safe idle condition, then set the second half by ear for minimum hum.

POP is from 1996 and I made a last minute change to the circuits in the Toroidal Output Transformers chapter, to include active bias balancing after testing some methods for this and from hearing about the importance of balance in a toroidal OT. Frankly, the change to POP was primarily to impress Menno vander Veen, who I was about to meet. I'm like everyone else and there are people I admire and am a fan of their work. Menno is a fantastically smart guy at the cutting edge of audio transformer science, and a totally lovable and down to earth guy to boot. In Menno's own book at the time, he used a faulty passive bias-set scheme that lacked range resistors and his kits followed suit. He did not bother with active bias until his later "trans" amp adventures which were mostly SE and often used EI OTs.

The "trans" amp is an idea of using local feedback around the power tube to essentially give the stage a virtual-earth input - one that take current in rather than voltage, so "trans" refers to transconductance, not a car transmission, or any of the modern identity liberation meanings Big Grin This idea is actually quite old, then revisited in the 1980s and then by Menno in the 2000s.

**Preferred bias circuits are ones that have a raw bias supply that is separate from the bias-set network. For example, the bias circuits for the projects in TUT3 or TUT5 are preferred. The usual bias circuit found in a Marshall is not.
Reply
#13
Thanks for detailed and informative reply! I find evolution of designer's ideas interesting, so it was really nice to hear the context for what is in POP. I've been thinking about trying a toroidal OT for a while and have been reading up on using them.
Reply


Forum Jump:

[-]
Come in where it's warm!
A warm welcome to tube amp modding fans and those interested in hi-fi audio! Readers of Kevin O'Connor's The Ultimate Tone (TUT) book series form a part of our population. Kevin O'Connor is the creator of the popular Power Scaling methodology for amplifiers.
Please remember these three principles: respect, sharing, community.
Not familiar with The Ultimate Tone book series? See discussion topics, or click here to visit London Power/Power Press Publishing.

[-]
Tube Amp Forum Hosted by London Power
London Power logo