Kits useable with a M500 300B Monoblock amp

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi,

One thing I do need to ask, I have studied the WE91 mod schematic and gone to the link you provided. I have read that info and downloaded the other appropriate mods.

However, the basic mod appears to be the introduction of a 10uf cap in front of the choke, with the movement of the 220uf caps to after.

This is nothing really to do with the "91" Design, it is a modification to the powersupply derived from my "Lady Day 91" Project, which is a slightly different but parallel take on the Lady Day 91.

I would suggest to use this in combination with the WE Style connection of the 300B Cathode shown earlier)

While this is clear, I have read comments about the "full" We91 mod.

The "Full WE91 Mod" being referred to is I believe based on this article:

Tube DIY Asylum: Oh no - YAWE91B300BA (Yet Another WE 91B 300B Amplifier) by Thorsten

This can hardly be called a modification, it basically requires the full stripping down of the chassis and a full rebuild to a new circuit.

The Diyhifisupply Lady Day 91 has been reviewed a few times, so you can refer to these reviews as to "what it sounds like" (and before you ask - the 6SJ7 is not a good substitute for a WE310A in my books).

DIY HiFi Supply - Ladyday Signature 300B SET 91 - The Ladyday and Billie shootout! - Review By Clive Meakins

Around the World: DIY Systems from the Other Side of the Globe

Something also possible useful may be this little article:

FAQ -- Which Amp Is Right For Me? | Diy HiFi Supply

I would say that the M500 with the hybrid Mu-Follower and gridchoke goes a long way to provide the same grip on speakers and the macro dynamics of the original Lux (the Lux 91 is another step up). I would probably add a 4.7K/4.7uF "snubber" (aka DR.P) across the coupling capacitor, this gives a little more dynamics in the lower ranges.

At the time when both the 6SL7 Mu-Follwer driven Lady Day and the Lady Day 91 where quite hot new items there where users that compared both and preferred one or the other. Those with very high efficiency speakers (Lowther Horns, big Tannoys, Bastanis Prometheus etc.) generally preferred the 91's while those with medium efficiency speakers (eg. AN-E's and the like) generally preferred the 6SL7 driven versions.

I hope this helps to make a call which way to go.

Ciao T
 
Thorsten, just to give some perspective, my speakers are very high effieciancy (102) speakers rated at 12 ohms, and they have a very flat impendance curve.
If that would lend a preference to certain mods, please let me know.
I am assuming that conversion of the amp I have to the 91 using the 310A is not practical.

As I will be making a lot of changes based on what we have talked about, plus recommendations from the links you have provided, I thought I would tweak the resistors to the highest grade that made sense, since there are so few.
I was going to switch them to the Japanese premium carbons that DIY carries. If you feel that the Tantalums would be better, let me know and I can switch or order both.

I am also assuming that the caps on the power supply board would not be an avenue to tweak, both in quality and in capacitance. In some of my earlier experiances (with much different tube amps), switching out the PS caps to higher quality, and higher value caps, led to a quieter and more stable/powerful supply.
Do you think that might be true for the M500?

Anyway, just to sum up: Based on your extremely kind suggestions and links, I will add the filament heater, power supply mod + WE cathode connection, Hybrid Mu,and a few other smaller cap changes.

Possible additions might be the resistors (since I'm already there) and the PS suppuly caps (again, I'm there and its not difficult if you think it makes any sense).

Since noise doesn't seem to be an issue, additional shielding doesn't seem needed and conversion of the 6SL7 to a 310A or a interstage transformer I assume is not possible/practical with this amp.

If this looks like a good understanding of what we've discussed, and you agree with my assumptions, I'm good to go.

If you have a chance, a comment on the resistor changeout and PS supply would be appreciated, however I want to sincerely thank you for all you have done. It has really been educational and I look forwards to the "new" M500's.

If I hadn't already gone this direction, the new amps would be high on my list and when these have some more hours on them, I suspect they will move to another room and I will be taking a very hard look at the latest and greatest. Its hard to beat great design and great prices at the same time!

Thank you very much, Bruce Prager
 
Hi,

Thorsten, just to give some perspective, my speakers are very high effieciancy (102) speakers rated at 12 ohms, and they have a very flat impendance curve.
If that would lend a preference to certain mods, please let me know.

Your speakers are very suited to a "91". However...

I am assuming that conversion of the amp I have to the 91 using the 310A is not practical.

It has been done, but it means stripping down the Amplifier almost completely and re-building it.

As I will be making a lot of changes based on what we have talked about, plus recommendations from the links you have provided, I thought I would tweak the resistors to the highest grade that made sense, since there are so few.

I would suggest to make the grid-stopper (in series with the grids) carbon composite (Allen Bradley, Xicon) or Takman Carbon if the others are unavailable.

As the Cathode resistors of the 6SL7 are bypassed their quality is not very critical, almost anything will work okay.

For the Gridleak resistors (from Grid to ground) I would suggest Takman Metal film, though again, this is not super critical. If you use the Gridchoke for the 300B this Gridleak resistor is omitted.

For the 47 Ohm in the Powersupply line (output stage WE connection) is best a 2W Kiwame.

If you do convert the Amplifiers to 91's then use Tantalums for the Anode Load on the 310A and on the 300B Grid Circuit, in that circuit these are quite critical sonically.

I am also assuming that the caps on the power supply board would not be an avenue to tweak, both in quality and in capacitance. In some of my earlier experiances (with much different tube amps), switching out the PS caps to higher quality, and higher value caps, led to a quieter and more stable/powerful supply.
Do you think that might be true for the M500?

You can just replace the PSU Cap's, sure. But since the demise of Black Gates there is very little in any format that fit the existing PSU Boards and wiring in capacitors with extra long leads into the existing PCB's (as often seen) is a really bad idea as it adds ton's of inductance over the original capacitors which generally makes at least the objective results (and to me also the subjective ones) much poorer than even cheap generic electrolytic capacitors.

However if you instead apply the PSU Modifications listed as "91" modifications and you apply the WE Connection to the 300B Stage using Film Capacitors (12uF & 47uF) you get a much bigger upgrade while reducing the impact of the powersupply much (to a larger degree than the PSU Cap changes could provide).

Anyway, just to sum up: Based on your extremely kind suggestions and links, I will add the filament heater, power supply mod + WE cathode connection, Hybrid Mu,and a few other smaller cap changes.

Yes, I think this will do best in terms of effort and money spend.

Ciao T
 
Thanks for all your help. I was wondering what advantage their is to using an interstage transformer in place of the driver tube. That seems to be one of the differences between
Audio Note 300B's and most others. I saw that DIY sells an interstage, but obviously felt using the 310 was better overall in their amp.
I would think the transformer would be more linear, however, I don't know about the sound, whether it would become a bit analyltical compared to an all tube design.
I was just wondering if that was a potential design that was looked at when you designed the latest Lux amps.
Thanks again and look forwards to getting started on my mods.

Bruce Prager
 
I have some wonderful scott amps that are el34 or el84. But I like 300B.

I have looked at the chinese 300B's, and have had a Cayin which was wonderful.

But due to economic conditions, I had to sell it and now am looking for a more reasonable amp. I found a used Billie, and am looking to optimize it as the total cost to optimize the amp, even to all of Thorstens suggestions, brings the total cost to way under the cayin amp. And the cheaper chinese amps have extremely mixed reviews.
So I don't think this is a crazy idea.

BP
 
Hi,

I was wondering what advantage their is to using an interstage transformer in place of the driver tube.

There is no advantage as it is not possible to do so. An interstage transformer essentially takes the place of the interstage coupling capacitor. You still need a driver tube however.

Most interstage transformers are 1:1 so they do not really offer any different signal as capacitor coupling. Of course, interstage transformers have a number of additional sources of non-linearity over a capacitor. Depending on how it is done the resulting Amplifier can be better or worse.

I have found that most interstage transformers are by far too limiting in bandwidth and have iron based cores (amorphous iron is still iron) which I dislike in transformers for low level signals. Tango and Tamura (James too IIRC) do offer some permalloy interstage transformers which I do happen to like, but a pair of these approaches "donating a kidney for cash" territory and still need some trickery in implementation.

I generally try to work without interstage transformers as a result.

The key difference between capacitor coupling and interstage transformer coupling is that the impedance at the grid of the output tube progressively increases as the frequency lowers, which means that at high power levels the grid current from the 300B is handled better.

My personal "budget" alternative to this is the use of a gridchoke together with "DR.P" (Damped Resonance Parallelfeed), which goes a long way to redress the situation or indeed "my latest trick" (Yes, that was Dire Straights on Brothers in Arms), which I call the "Bushido" circuit. It forms the core of diyhifisupply's lates generation of 300B Amplifiers, the Lux and Lux 91 but cannot easily be applied to your Amp's.

That seems to be one of the differences between Audio Note 300B's and most others.

No, not really. Very few AN-UK Amp's ever used interstage coupling and AN-JP never used any IT coupling. In fact, AFAIK non of the AN-UK 300B AMp's use interstage transformer coupling and only the AN-UK Ongaku and Ankoru use IT coupling.

Here a schematic of the quintessential AN-JP Amp, the Ongaku:

Audio Note Japan (Kondo) Ongaku

I would think the transformer would be more linear

Why would you think that?

[/QUOTE]I was just wondering if that was a potential design that was looked at when you designed the latest Lux amps.[/QUOTE]

Above I noted that there where/are significantly different sonic traits for the Lady Day in the 91 (WE 310A Pentode driven) and standard (6SL7 hybrid Mu-Follower with Gridchoke) configuration.

The 91 Style Amplifiers I have worked with all excel in delicacy and detail as well as in "tone", by far more so than any others (this parallels the Yamamoto Amplifiers with their Pentode drivers), however they start compressing and clipping fairly early.

The Amplifiers with low impedance triode drive (be it using low impedance triodes with or without interstage transformers or using such circuits as the hybrid Mu-Follower) always seem to offer a more robust presentation, they do not shrink from big orchestral climaxes and handle big bomblast Rock well, but they do not seem to offer quite the same redition of detail, space and tone.

Of course, ideally we want a combination of both. The two Lux amplifiers attempt to deliver exactly that and manages quite well I think.

If you read the review in 6moons, the Lux 91 went up against the Border Patrol 300B Amp's, which are IT coupled and are designed by a guy from the general original audio innovation clique - which later also gave rise to Audio Note UK and Audion UK among others.

My "Bushido" circuit performs both subjectively and objectively much better than any IT I ever tried in terms of linearity, bandwidth and ability to handle grid current and it frees us up to use any voltage gain stage we like, including the 310A.

However, this circuit and design is not applicable to your Amp's (sorry), you don't have enough tube sockets and space inside etc.

Ciao T
 
Interstage or not to interstage...........

Thanks for the explanation. I asked as on the Audio Note kit page, they have a mono block kit that uses an IST and suggest that it has advantages over a non-IST design.
As you say, you can design amps many many ways, and you get advantages and disadvantages, as well as prices differences that can many times explain the advantages very quickly (the "you don't really need any kidney's pricing").

I'm planning on the MU circuit. The DR.P mod you mention, is that part of what I'm doing already with the Gridchoke mod or is it a additional mod for me to consider?

Not a big deal either way, but my philosophy is that once I'm in there and its not going to cost me a body part, I may as well do all I can so that I can mod once and enjoy forever.

My other comments regarding IST's basically came from what was on that page for the amp (I think they call it the Mono interstage if I remember correctly). My thought on the linearity was based on the premise that you would not be driving the IST near saturation levels and that a transformer (well-made) would not be subject to the sound differences that occur with caps due to the large number of different materials and designs that occur with them.

Of course, that is also why I wondered about it being a bit analytical, as those difference in caps can be judiciously used to tune a amp to the sound you prefer. I believe in neutral amps, but am not one who believes any departure is a flaw. Rooms and ears are anything but neutral and for me, the idea is that the music should be enjoyable. If its not "accurate", people have been arguing what that word even means even today.

I went with the Billie because I have developed a real love of 300B amps, seconded by small tube amps like the Scott 222 series. I prefer the slight added warmth in the mids (so long as it doesn't get syrupy) and the Billies have that quality. The only issue I have had is that the highs are sometimes a bit hot with my speakers, (which is why I'm going with the AN copper coupling caps) and hope the mods will tame that issue without destroying the underlying detail, and keep the wonderful mids I love.

Much of my experience in modding or maintaining amps have been with the Scotts which are very different than the Billie's, particularly as the Scotts are integrated's.
Its actually nice to be working on a circuit where you have some room to work and tracing the schematic is so much easier.

I really appreciate the detail in your responses as they have given me a much better understanding of 300B amps like the Billie, which will be highly useful in both modding and maintaining it as I intend to have it for some time to come.

Best regards, Bruce Prager
 

Attachments

  • Mono_Schematic003.jpg
    Mono_Schematic003.jpg
    56.8 KB · Views: 720
Hi,

Thanks for the explanation. I asked as on the Audio Note kit page, they have a mono block kit that uses an IST and suggest that it has advantages over a non-IST design.

I do not think it would be approriate to make specific comments about specific products from a competing manufacturer here. But I would look at what AN-UK and AN-JP do in their best 300B Amplifiers...

I'm planning on the MU circuit. The DR.P mod you mention, is that part of what I'm doing already with the Gridchoke mod or is it a additional mod for me to consider?

DR.P means you connect a snubber (I would suggest 3.3K & 3.3uF) across the coupling capacitor (I would use 0.33uF or 0.47uF with the above snubber).

My other comments regarding IST's basically came from what was on that page for the amp (I think they call it the Mono interstage if I remember correctly). My thought on the linearity was based on the premise that you would not be driving the IST near saturation levels and that a transformer (well-made) would not be subject to the sound differences that occur with caps due to the large number of different materials and designs that occur with them.

Hmmm, transformers have many problems of their own. airgapped transformers have very limited primary inductance, which limits their LF response, also, air-gapped transformers rarely use permalloy cores, so their sonic merit is quite limited simply by the core material (worse if they do not use 0.05mm steel laminations).

Equally, they need to be quite large physically (to handle substantial DC current) and need high impedances, which means winding for good HF response is at the best very difficult. Of course, few transformer winders attempt to make IT's that offer the same wide bandwidth as capacitor coupling routinely allows. Due to the "law of the 400k" this tends to work okay sonically, even though the bandwidth is obviously limited, if the transformer is designed well.

Compared to for example to a CCS Load with a gridchoke an interstage transformer will usually have much worse (lower) impedance at low and high frequencies, only in the midrange can it offer similar linearity.

In addition, there is usually a claim that interstage transformers removes coupling capacitors. This is of course patently untrue, refer to Lynn Olson's "Current Loop" presentation at ETF (available at nutshellhifi.com) for more.

So, in the end IT coupling can have some advantages over capacitor couplings, but most of these can be attained using other means as well.

Making a good Interstage transformer costs as much if not more than making a good medium power output transformer and even then the transformer is unlikely to match the bandwidth of other options. If this is a possible problem or not depends on musical tatse and speakers used (e.g. those with Speakers that omit low bass and high treble will not notice the problems, other very well might).

hope the mods will tame that issue without destroying the underlying detail, and keep the wonderful mids I love.

The "Hot high frequencies" are mostly (IMHO) down to tube choices and the heater supply. Try some good NOS 6SL7's and warm or neutral sounding 300B's. If the Amp has it's original tubes, these sound quite steely and bright.

With my Billy Kit I did not listen to the supplied Chiense tubes (O&J aka Valve Art) for long, the chinese 6SL7's where pulled within twenty minutes of listening. I used with the Stock Billy Svetlana 300B's (sounded quite close to WE) and some nice NOS 6SL7's, I liked the Ken Rad ones very well. I also rather liked the TJ Tubes.

The Heater supply modification tend to reduce edginess and grainyness.

The Powersupply Mods and WE connection of the output stage extend this, the WE connection also gives a more solid presentation and more of the oh so famous "black background".

The Hybrid Mu-Follower adds more detail and slam.

Gridchoke and DR.P extend slam and make the Amp subjectivealy capable of playing notably before compressing or breaking up.

Ciao T
 
Thanks as always for an informative answer. So many sites list the positive, but not the potential negatives, its nice to get a balanced presentation. At least in my case, I learn a lot more.

Thanks a lot for your thoughts and info on the high end issue. Like you, I yanked the stock 300's very quickly and substituted Princess mesh tubes that I have liked in the past. They significantly helped in all areas.

I also am playing with both the rectifier and driver tubes. Changing to a GZ34 made a surprising change in the sound, cleaner with better mids. It was a large enough change even my wife noticed it immediately and I swapped again just to make sure I really was hearing a real change. So those are staying in, and the last experiments in tubes, are for the 6SL7. A friend is letting me borrow some ECC85 NOS mullards, a set of 5691's and a NOS 6SL7 whose brand escapes me at the moment.
I figure thats enough to see where to go for now, and then the mods should simply help out the amp overall.

I haven't had a chance to try the TJ tubes and if I get the chance will see if anyone I know have a pair. I have heard good things about the brand in general.

Thanks for your help and all the great info. I appreciate the time you've taken to answer all my questions.

Best regards, Bruce Prager
 
Hi,

I also am playing with both the rectifier and driver tubes. Changing to a GZ34 made a surprising change in the sound, cleaner with better mids.

Try some coke bottle, brown base 5R4GY's. I liked them best.

the last experiments in tubes, are for the 6SL7. A friend is letting me borrow some ECC85 NOS mullards, a set of 5691's and a NOS 6SL7 whose brand escapes me at the moment.

Try finding Ken Rad's (VT-229 is the type designation if I remember correctly). I know they are priced at ridiculous levels, but they live up to the Hype.

I haven't had a chance to try the TJ tubes

It is my understanding that the Sophia Tubes come from the factory and are made to identical spec's as those sold under the TJ brand. Of course, there may be differences in selection and warranty etc. that make one a better than the other, but if you have one I doubt you need the other.

Ciao T
 
Thanks, I'll give them a try. Hadn't heard of them. I looked for Ken Rads but so far the only ones I could find cost more than the amp! But I'll keep an eye out.

I didn't know that the Sophias and TJ's were "related", but it makes sense. The sound of the Sophias are just about what people write about with the TJ's.

Thanks!!

Bruce
 
Hi all, especially to Thorsten,

I haven been planning some mod to my old Consonance Reference 5.0 integrated amp which I think is very similar to M500 the main superficial difference would be the driver tubes. On Reference the input is 12at7, and driver is 12Bh7.

Upgrade i have planned for this amp includes:

1) Power Supply: changing two in-series Rubycon 220uf/450v to one Jensen 100uf/550v electrolytes. I wish I could squeeze in the mundorf tubecap (film cap) but they are expensive and size are too big to fit under the chasis. I will bypass both 100uf with a small value film cap of 0.1uf/630v solen.

2) Cathode Bypass Capacitors: I wish I could replace all of them by film caps but again size is a big concern. For input and driver tube, there are 100uf/16v BG PK for each, I plan to change to the red Blackgate NH series. Cathode bypass on the 300b's 1kohm Dale resistor is a rubycon 220uf/250v electrolytes. This is the main area giving me problem and need some advise here:

a) I want to replace it with film caps, but 220uf film are way too big, i cant' find space in the bottom to fit all these in. 2 x 100uf films are even bigger. Can I go down to 100uf (Axon truecap) ? what would I lose?

b) And a bit more extreme, can I do away the bypass cap completely, and reduce the resistor from 1k to 470ohm to bring back the gain lost in going to unbypass cap? Is this mod worth or it has major flaw?

3) original coupling caps from factory are 0.1uf and 0.47uf Mundorf supreme. I have got some russian ksg silver mica of 0.1uf on hand and planning to replace the 0.1uf Mcap by the silver mica 0.1uf and for the 0.47 Mundorf, i just add the 0.1uf silver mica on top. I have some russian teflon ft-3 but can't decide if i should put in teflon or mica for a while, eventually I settle on the silver mica.

4) for the filament heater supply, originally is rubycon 10000uf/16v, I have some Jensen 10,000uf/65v on hand from past projects and plan to swap these in.

Need help here, and thanks in advance

BTW, my speakers are Bastanis Atlas with gemini dipole tweeter and 18" dipole mandala bass bin

I am from Hong Kong
 
Last edited:
Hi,

I haven been planning some mod to my old Consonance Reference 5.0 integrated amp which I think is very similar to M500 the main superficial difference would be the driver tubes.

Sorry, I have no experience with this Amplifier.

1) Power Supply: changing two in-series Rubycon 220uf/450v to one Jensen 100uf/550v electrolytes.

I would check other options. I am unsure you would realise a major upgrade from this, instead of first taking care of other areas and fundamental design flaws.

I will bypass both 100uf with a small value film cap of 0.1uf/630v solen.

I'd be really careful on that. Better use a small value resistor between the electrolytics and the bypass cap and increase the bypass cap as much as possible in value.

The ESR of the electrolytics will be in the region of 1 Ohm or less, meaning an 0.1uF capacitor would only be effective above several MHz, where most axial film types are not very efficient, in fact, most likely you will get several tank resonance circuits above a few 100KHz that make a real mess up there, while not really helping anything in the audio range.

If instead you inserted 33Ohm into the PSU Line and connected a 10uF capacitor after this resistor you would have a system that is effective from 500Hz upwards and can effectively reduce high frequency noise from the supply and likely will give a much greater benefit than swapping out the Rubicon's for more of the same with a different label on it and would reliably have non of the issues your bypassing may (will?) create.

2) Cathode Bypass Capacitors: I wish I could replace all of them by film caps but again size is a big concern. For input and driver tube, there are 100uf/16v BG PK for each, I plan to change to the red Blackgate NH series.

I think (depends on the exact circuit) you can omit one or several of these. The input most likely without second thought, the driver will depend on the exact circuit. If you look at the Ongaku circuit for example, you find only one cathode bypass cap which is in series with an unbypassed 1K resistor, so 33uF good quality film will do nicely there. The alternative would be to drop the DC coupling and use an interstage coupling cap, which would probably a better choice, all considered.

Cathode bypass on the 300b's 1kohm Dale resistor is a rubycon 220uf/250v electrolytes. This is the main area giving me problem and need some advise here:

a) I want to replace it with film caps, but 220uf film are way too big, i cant' find space in the bottom to fit all these in. 2 x 100uf films are even bigger. Can I go down to 100uf (Axon truecap) ? what would I lose?

As a fundamental thing, in many circuits the Cathode bypass capacitor values are very much oversized. With generic electrolytic cap's this can help.

In one of his WW article's arch-objectivist Douglas Self compared (measurements) "audiophile" capacitors to the same value generic types (Audiophile types in my reading are cap's designed for good audio performance, which is met only by few of those commonly marketed as audiophile). The audiophile cap had much lower distortion. In addition D. Self found that increasing the value of the generic cap reduced distortion to comparable levels.

I find that often 22uF Film Capacitance can replace 220uF of good quality generic electrolytic capacitors as long as this does not cause problems with low frequency response. For a 300B with a 1K cathode resistor I personally find 33uF the lower limit. Any less and you measurably loose ultra low bass (even if not audible with most speakers).

You can instead of the 220uF electrolytic cap use my style of WE connection, with a 12..15uF/630V film capacitor added from the top (+B) of the output transformer to the 300B Cathode and the cathode capacitor changed to 47uF/160V (to keep size small).

Due to the way the WE connection works, even an electrolytic capacitor is acceptable from the cathode to ground, though a film type is still giving a modest improvement over an electrolytic cap there, but nothing like the kick just the WE Cap alone brings.

If you have time, tuning the WE Cap for best power supply rejection can give a much greater benefit than large amounts of money put into expensive PSU Cap's in a conventional circuit. Indeed, I'd go as far as saying that generic PSU Cap's with a well implemented WE connection sounds better than a generic circuit with monster size super premium film cap's replacing the original cap's and the one on the 300B cathode.

b) And a bit more extreme, can I do away the bypass cap completely, and reduce the resistor from 1k to 470ohm to bring back the gain lost in going to unbypass cap? Is this mod worth or it has major flaw?

Well, first, if you change the 300B cathode resistor value you change the current in the tube, going to 470 Ohm will essentially nearly double the current, which will fry your outputs crispy quite quickly.

Secondly, the bypassing of the output tube cathode resistor makes sure it has a low effective anode impedance. With a sufficient value bypass capacitor you in essence get the tubes anode impedance.

A simple way of guesstimating the effect of unbypassing a cathode resistor with a triode is to take the value of the cathode resistor, multiply it with the gain (Mu) of the tube and add this resistance to the Tubes own anode impedance.

So, a straight 300B will have around 700 Ohm Anode impedance (and a gain of 4), with a cathode bypass capacitor this will be maintained at all frequencies but very low ones. The output transformer is normally 3K primary, so your effective gain from the tube is 3.2. The Transformer will turn the 700 Ohm anode impedance of the 300B to a theoretical 1.8 Ohm, to which we normally need to add some output transformer losses, which gives the commonly found output impedance of 3 Ohm for the 8 Ohm tap of a "common garden" 300B Amplifier.

If we now have a 470 Ohm cathode resistor (and ignore at the moment the smell of frying expensive 300B Tubes) this resistor will appear now as an effective addition of 1.9 KOhm to the Anode Impedance of our 300B. As it is now run well past official limits our anode impedance should be around 450 Ohm (more current = lower impedance), so in total our output imepdance is now around 2.3KOhm. This means our output impedance from the 8 Ohm tap more than doubles and our gain is down to around 2.2.

So, I think you will find that doing what you suggest is likely a very bad idea on several accounts.

You could convert the amp to fixed bias, but then you need to consider the quality of the final power supply capacitor again very seriously and you would have to probably completely redesign the power supply and you may even need a completely different power transformer. Still, could be worth it.

In your case I think WE connecting the output stage will be the easiest, quickest and highest impact mod, next to properly fixing the 300B heater supply. Next to these two the things you have planned will be mere small change in sonic terms (if not in cost).

3) original coupling caps from factory are 0.1uf and 0.47uf Mundorf supreme. I have got some russian ksg silver mica of 0.1uf on hand and planning to replace the 0.1uf Mcap by the silver mica 0.1uf and for the 0.47 Mundorf, i just add the 0.1uf silver mica on top. I have some russian teflon ft-3 but can't decide if i should put in teflon or mica for a while, eventually I settle on the silver mica.

Try the various capacitors, you may also find that 0.1uF suffice for the coupling to the 300B, depending on the precise design/schematic of the Amp.

4) for the filament heater supply, originally is rubycon 10000uf/16v, I have some Jensen 10,000uf/65v on hand from past projects and plan to swap these in.

This may bring a minimal improvement at the expense of a huge lump of cap, but fails to address the fundamental problems. Using Schottky rectifiers instead of the original bridge and a pair of 2,200uF caps with a pair of around 0.47 Ohm resistors between them can bring a by far greater improvement across the board and keeping the 10,000uF rubicon as second cap with 2,200uF first cap and schottky rectifiers will produce even greater gains.

Need help here, and thanks in advance

I suggest you first clear your head of the accumulated audiophile ** (e.g. the concepts that component swapping without addressing fundamental design flaws can magically transform gear, that most high end designers actually know what they do and their circuits are correct and that the values of a given component is optimal the circuit and you must always replace like for like on values and and ratings, even if it means the much large component is stuck somewhere miles from where it belongs and connected with long wires and so on...).

Then get a circuit diagram, a scope and some basic test gear and find the actual problems, having found them, resolve or at least improve them.

This will allow you to indeed "transform" the gear, but not by magic, but by sound engineering (pun intended).

Don't get me wrong, parts quality does matter quite a lot, but a good design sounds very good with generic parts and even better with excellent ones in the right places, a poor design still sounds poor after thousands of dollars worth of magical caps and so on thrown at it.

Ciao T

PS, I am often in HK on business, if you like, we can meet for yum cha and I can explain many of the principles much better with pen and paper than here.
 
Coupling cap changeout

Thorsten, received everything I think I need, and have mapped out the mods, but the changeout of the coupling cap has me a bit confused.

I am planning on two changeouts based on our discussion.
Looking at the schematic for the M500 you posted, the cap labeled C1 100uf is being changed out to 22uf film caps as per our discussion.

The other coupling cap you refer to I assume is C3? which is 150uf on the schematic.
The confusion is that the link I ordered from specified the new cap as a .22uf cap, which I choose as a Audio note copper PIO.
Am I really changing the 150 to a .22, or do I have the wrong cap (and/or ordered the wrong value cap)?

Also, to make sure I'm in sync with the schematic, the grid choke replaces R6 on the schematic?

Everything else makes sense and I have determined the best way to install them, its just the value of the cap that has me thrown a bit.

Thanks for all your help!!

Best, Bruce Prager
 
Last edited:
Hi,

Thorsten, received everything I think I need, and have mapped out the mods, but the changeout of the coupling cap has me a bit confused.

I am attaching a schematic of the M500 with my suggested MINIMUM modifications applied. This does not show the Filament Supply, but a simple CRC filter, for thse who want to keep costs to a minimum.

The coupling capacitor is marked C2 on both Schematics. Value is uncritical, minimum 0.1uF. My mod schematic shows a 0.47uF coupling capacitor. The 4.7uF/4.7K "snubber" across the 0.47uF cap is the fabled DR.P.

In simulations I have found that making the DR.P cap 10 times the value of the coupling cap and making the resistor in KOhm equal to the DR.P cap in uF seems to work best, but I have not done much experimentation. So feel free to try other values etc.

Also, to make sure I'm in sync with the schematic, the grid choke replaces R6 on the schematic?

No, R4 (the grid leak resistor).

Also, my schematic shows the original power supply, however I would recommend to convert it to the schematic shown in post #2.

Ciao T
 

Attachments

  • M500-Mod.jpg
    M500-Mod.jpg
    562.2 KB · Views: 561
Thanks Thorsten.
The pwr supply change is the addition of the 10uf cap before the choke and the move of the 220's? I have that mapped and have the 10uf as a gold Obligatto.
Also, I am adding the 47R resistor and the 22uf cap as was circled in green on the first diagram in post #2.
Didn't realize quite how big that 22uf bugger was going to be however extending the case depth is quite easy so I'll stuff it in somewhere. I notice similar ones are used your new amps. Those are really packed, I definitely admire the skill it took to lay out that much componentry and get it all to fit so well.

I originally ordered .22uf coupling caps. Can/should I still use these, or return them and get .47's? I had picked the copper AN PIO's.

The 15uf and snubber components? Do they need very high quality components or are they not as critical? I could go with Hovland or Auricaps if critical, resistors aren't hard to get.

Everything else seems to match my workplan, so I really appreciate you taking the time for the schematic. It helps to have a frame of reference when you are doing things like this. Otherwise its easy to get lost and misunderstand what goes where.
This was a really big help.


Thanks again for your help and support.

Best, Bruce Prager
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.