SKYLINER build - inspired by ODS #124 and #060

[Stephen1966]

By consistency on the dropping string, I believe Ian and I are referring to the 22k and 2k2 from PI to OD and OD to clean, which produces a more or less standard ~300V on the PI plates, ~200V on the OD, and ~190 on the clean.

Thanks Martin, good point. In my own ham-fisted way this is what I was trying to say as well. The current PSU isn't giving me the voltages I need yet. But I believe these voltages will bring the amp to life. It's at the top of my list, systemic problems aside, this one may be the most important.

[ijedouglas]

The current PSU isn't giving me the voltages I need yet. But I believe these voltages will bring the amp to life. It's at the top of my list, systemic problems aside, this one may be the most important.

If your voltages are in the correct ratio but just lower than you want, you can raise them with a variac to see if the increased voltage gives you what you are looking for.

[Stephen1966]

The current PSU isn't giving me the voltages I need yet. But I believe these voltages will bring the amp to life. It's at the top of my list, systemic problems aside, this one may be the most important.

If your voltages are in the correct ratio but just lower than you want, you can raise them with a variac to see if the increased voltage gives you what you are looking for.

Good practical advice. Thank you.

[Stephen1966]

...I decided to really push it in OD. Again, this was with R/J switch in the PAB position and here, I get feedback more readily. I also get the "sizzle" on the bass notes particularly, less well defined on the treble side. And, in the higher positions I get that kind of hollow tubescreamer kind of effect (don't know what else to call it) on the melodic notes. If I put it in Rock mode, the effect is muted and in Jazz, it disappears altogether. This is really loud though, I have a concrete floor in the basement and I can feel the sound vibrations in the soles of my feet. Stuff was rattling all over the shop. Fortunately, none of that rattling seems to coming from the cab. After about an hour of this, I switched it back into clean mode and instantly felt the volume and sound pressure drop away. Like half volume, even though I didn't touch the Volume or Gain. The Level and Ratio of the OD was about 45% and 55% respectively, at other positions, I lost the tubescreamer effect. The purest high notes though have a slight, unpleasant distorted sound at the edge.

Thanks for the recent posts, as I mentioned to Tony, these conversations are really helping to put things in perspective. I see a number of tweaks I should perform and there isn't any single thing that will get me all the way there, though I think this one, to do with voltages and gain, is a biggie!

I was thinking about that "distorted sound at the edge"... I am quietly convinced this is clipping. I observed it when I was scoping out the balance of the PI tube, turn up the Master high enough and sure enough it starts to clip. This was at about the same level of gain, yesterday. Reducing the gain diminishes the effect of bloom and it is all but absent in the clean channel - a slight fattening of the sound perhaps. So it's clear that gain pushes the effect but the problem is, at the moment, it all comes from the Master, and ideally it should emerge sooner in the preamp, then I might dial the master back and avoid the clipping. I think it's a good idea to play with the variac as a diagnostic tool for looking at the proportion of voltages across the dropping string. I will still need to reduce the resistor feeding the PI node to bring it up to the ~300V region. I can't remember the exact figures now, but I think there is already a 10V difference or thereabouts between the OD and Clean, that may change proportionately as the PI is increased but if I still get the 10V difference with the PI at 300V, just not at the 200/190 but somewhere close, I can look at the resistor feeding the OD at B+7, or if the reverb is on the low side, the resistor feeding B+5 and then B+7... I'm sure you get the picture. I firmly believe as well, that MrD's voltages are no accident, and that the target they represent is instrumental in bring the circuit to life. Making it not just a nice amp. But a great amp.

An additional tweak I will be implementing as well is to swap out the 120k/5k/110k resistors on the PI plates for something more along the lines of the #183: 110k/10k/100k. The cathode resistor and tail are the same in the #124 and #183 so I know I'm on safe ground there. It should also give me a little extra voltage to play with and a lot more ability to control the balance than is offered by the 5k trimmer. The prospect that it might give me a bit more bite in terms of the sound is perfect, but I now know better than to expect a huge difference from small changes like these.

There's more... like looking at the effect of changes to the LNFB and GNFB as well. The heart of the amp is the #124 though and these nets probably don't figure too highly in the tweaks when their current values are inherent to the design. They may even be necessary to the "signature sound", so its a case of all things in perspective - systemic thinking. You've gotta love it!

I really love the complexity of this amp and as frustrating as it might seem when I throw the switch and it doesn't perform perfectly, as expected, working through the problems and understanding the complex relationship between the elements isn't just fun, its a valuable learning experience as well. I'm sure I'm going to get there eventually, and when I do I will look back at this process I will be able to say I am glad it wasn't perfect at the beginning. Anyone still wondering why it took MrD five years to build an amp though?

So now, I have to order some more parts and wait for the next free day when I can get back into it, I have some time on Wednesday morning though so I will have a go with the variac then. Probably next weekend if I can get my orders in today...

Speaking of which, does anyone know of any European suppliers for CTS j-taper pots? I should check Mouser as well but when I went looking for these before I couldn't find anyone outside the States who had them.

[Edit, neither Amplified Parts nor Antique Electronics Supply will ship orders less than 150 EUR.]

[Richard1001]

Here you can get some vintage taper / 30% pots

https://www.tubeampdoctor.com/ersatztei ... 5&n=24&s=7

[Stephen1966]

Here you can get some vintage taper / 30% pots

https://www.tubeampdoctor.com/ersatztei ... 5&n=24&s=7

Many thanks, Richard. Ordered and on their way

[Stephen1966]

Here is a chart I compiled after trying Ian's suggestion of turning up the variac to see if I could get the effect I was looking for - bloom/feedback etc.

Voltages 24-11-2021.jpg

It certainly improved the sound and tone. I tried (I think) Henry's suggesting in the "How to set the PI" thread of listening to the bloom on notes D, D#, E on the second string. The D# was particularly lively, and started to feedback in OD with the mid switched in and the R/J switch in its PAB position. Emaj chord, nice feedback, Amaj. and so on, the beefy chords.

I didn't want to hammer it with the variac at close to 264V though so I didn't exactly put it through its paces. An improvement yes, and with further adjustments to PI, I think I could access that feedback more evenly across the notes. But altogether a pretty grissly tone (depth and fullness aside) and no surprises when the power section was being driven so hard. I didn't rebias at the hotter voltage, I just wanted to see what the voltages across the board would look like when I set the PI to around 300V.

The R values in my leftmost column are my current resistor values compared with the #124 and the #060. The #060 is a bit of enigma, the sources for these came from the hand drawn layout, the notes Chris Barnett made and the photos.

[to be continued...]

[Stephen1966]

For reasons best known to himself, when Chris Barnett was documenting the #060, he didn't like the reverb and seems to have decided at the outset not to bother documenting it but there does appear to be one observation of the plate voltage on V5 which is responsible for the input and output of the reverb. As if he was recording the voltages but stopped when he realised it belonged to the reverb. Who knows? I think his comment was something along the lines that he "could do better". It's a shame because it's left us with an incomplete picture of the amps topolgy and left a great deal to trial and error. One interesting observation that is clear though, is that MrD had set the Clean tube plate voltage at around 200V and the OD tube at around 210V, which is hotter than the #124 ODS by 10V and 10V respectively. The 7025s and 12AX7s will take this comfortably, I'm just wondering if that was done purposefully, to compensate for the 6dB (rootz's simulation) that the reverb circuit loses on the signal from the preamp? Afterwards though, the PI runs around 300V. The #060 resistor values are from the hand-drawn layout.

You will see from the chart that my voltages are a bit all over the place once I crank up the variac and I guess that was to be expected because the valves themselves are behaving like non-linear resistors and imbalances in loading will be exacerbated by the imbalances further up and down the line. It's given me a good understanding of the current state of my amp though, and ideas for the direction I need to take, with resistor values and the next steps in the process.

I have a simple question for you. My first step will be to adjust the PI plate resistors. To 110k/10k/100k. That will allow a little more voltage through the plates. The way I imagine the next step is to begin with (B+4) the PI resistor (the big 5W cermet in my build) and adjust this down to hit 300V at the PI, and then to address the next in the filtering stages B+5, B+6 and so on down the line. Should I expect any significant change in voltage at the PI when I start adjusting the resistors down the line? And similarly for the next stage etc etc?

I think I can interpolate the voltages of the reverb stages from the resistors documented in the PSU of the #060. the B+3 branches off the main filter string so can be read as it already is with the 22k resistor. It may be as easy as to find the correct value of the B+4 (which gives 300V at the PI) and then just replace the remaining resistors in accordance with the information we have for the #060. Anyway, I feel confident that adjusting the PSU to match the plate voltages is going to give me a much better, cleaner tone and I just have to wait for the extra parts before I can start on it.

Thanks to the generous support of Richard as well, I will be able to rationalise the response of the Volume and Master to a level of sanity with the j-taper pots which I believe will be arriving tomorrow. This is lower priority though, so my first task will be to modify the PI plate resistors and trimmer and then, the PSU dropping resistors.

It will be my first time doing anything like this though, so if I have overlooked anything, please chime in.

[ijedouglas]

I went back to look at your gut shots and noted a couple of things you may want to try:
- #124 had Dale RN65s on the plates, not sure what you have on your plates, but I would definitely swap those out. New production are pretty close and will get you there. (there was a huge debate on this and I don't want to go there. If you can find the NOS version...awesome....otherwise new production are fine)
- Lead dress on V1 - Notice how Mr Dumble tended to run V1a cathode parallel and touching V2b plate wire. This I believe was intentional and has been discussed before on the forum. Pay special attention to the lead dress and try and replicate it.
- GNFB resistor - looks like you have a red metal oxide. Try and see if you can get a 1 watt Piher like in 124
- slope resistor was a Roederstein MK3 150K in 124. Not sure what you are using there
- Pre-amp tubes - this has been discussed before on the forum. Work on getting V1 sounding good and work your way back till the PI tube.
- Reservoir caps are 220uf. 124 used 300uf. In my amp I used 330uf measuring around 310uf, Increasing this will stiffen up the bottom end a little. You may want to try this although it is a much large modification.

I would try all of this first before changing your dropping string. Like I said before, I had more luck changes voltages with tubes than by tweaking droppers.

PM me

[talbany]

I have a simple question for you. My first step will be to adjust the PI plate resistors. To 110k/10k/100k. That will allow a little more voltage through the plates. The way I imagine the next step is to begin with (B+4) the PI resistor (the big 5W cermet in my build) and adjust this down to hit 300V at the PI, and then to address the next in the filtering stages B+5, B+6 and so on down the line. Should I expect any significant change in voltage at the PI when I start adjusting the resistors down the line? And similarly for the next stage etc etc?

It's back
Depends? I went through this when I built my Music Man P.S. So the drill is? If your V1 and V2 are in spec and your PI is low then your going to have to change the dropper feeding the P.I as well as the dropper feeding V2 (OD tube) The last one (2k2) is usually pretty close. Depending on where you branched off for the reverb you may also have to change that value as well (as long as the rest of the tubes are in spec).
Remember preamp tubes are cathode biased so depending on the current draw (transconductance) of that particular tube, set at the given RK value at a given voltage you could get up to a 10V swing after the tube is biased and conducting, it's the same thing with cathode biased power tubes. in short? it's actually the current draw that sets the voltage. IMO a lot of players can get confused when they pop in a different tube and think it's just the brand of tube that's causing the shift in tone when it's really a shift in bias and plate voltage and I am sure the type of tube also plays a role as well, so really it's both
BTW.. I would bet anything Dumble was pretty selective about which tubes he used (including balanced triodes) in his amps.(Details)

The Pots.
A 2-35 CTS J- taper Dumble used can range anywhere from 30-35% taper. This means that with the pot set halfway (12 O:clock) a 1M J taper will get you between 30-35K. An alpha or 10K taper at noon you will be at 10k 1/3rd the value of the CTS.. So when players look at how others set their Dumbles and try to set them the same the amp performs completely different, Yes I am sure it does, this is really what gives the J-tapers their sound because they react more like a linear pot than an audio.

Tony

[Stephen1966]

Thank you Ian, a lot to digest here, and my brain shut down about two hours ago... I'll be able to think straight when I've had some sleep.

To be getting on with here are the components I used that you weren't sure about (with Mouser numbers).

- #124 had Dale RN65s on the plates, not sure what you have on your plates, but I would definitely swap those out. New production are pretty close and will get you there. (there was a huge debate on this and I don't want to go there. If you can find the NOS version...awesome....otherwise new production are fine)

594-MBE04140C1003FC1 Vishay/Beyschlag, MF 1W/1%

- GNFB resistor - looks like you have a red metal oxide. Try and see if you can get a 1 watt Piher like in 124

594-5093NW4K700J Vishay / BC Components, MF 3W/5%

- slope resistor was a Roederstein MK3 150K in 124. Not sure what you are using there

594-MBE04140C1503FC1 Vishay/Beyschlag, MF 1W/1%

I'll be in touch. A most generous offer, thank you.

And to all my American friends out there - Happy Thanksgiving.

[Stephen1966]

It's back
Depends? I went through this when I built my Music Man P.S. So the drill is? If your V1 and V2 are in spec and your PI is low then your going to have to change the dropper feeding the P.I as well as the dropper feeding V2 (OD tube) The last one (2k2) is usually pretty close. Depending on where you branched off for the reverb you may also have to change that value as well (as long as the rest of the tubes are in spec).
Remember preamp tubes are cathode biased so depending on the current draw (transconductance) of that particular tube, set at the given RK value at a given voltage you could get up to a 10V swing after the tube is biased and conducting, it's the same thing with cathode biased power tubes. in short? it's actually the current draw that sets the voltage. IMO a lot of players can get confused when they pop in a different tube and think it's just the brand of tube that's causing the shift in tone when it's really a shift in bias and plate voltage and I am sure the type of tube also plays a role as well, so really it's both
BTW.. I would bet anything Dumble was pretty selective about which tubes he used (including balanced triodes) in his amps.(Details)

The Pots.
A 2-35 CTS J- taper Dumble used can range anywhere from 30-35% taper. This means that with the pot set halfway (12 O:clock) a 1M J taper will get you between 30-35K. An alpha or 10K taper at noon you will be at 10k 1/3rd the value of the CTS.. So when players look at how others set their Dumbles and try to set them the same the amp performs completely different, Yes I am sure it does, this is really what gives the J-tapers their sound because they react more like a linear pot than an audio.

Tony

It's back and so are you Well, perhaps not today...

Talking about the current draw of a tube, a good example of that was the voltage swing I saw in the PI tubes. The first an unbalanced 7025 gave a 10V difference between the triodes, the next a balanced 12AX7 gave just about 1V difference. I know they are different, but also essentially the same tubes with a similar mu.

When you mention how players set their Dumbles, I was reminded of an interview with Robben Ford talking about the settings on his amp. I think, subconsciously at least, I was making a comparison between his settings and mine (thinking mine was somehow defective which is always possible) but it occurred to me that his probably has the j-taper pots installed. So what you're saying makes absolute sense.

The mouser order arrived today, so I'll be working on the dropping string this weekend.

In the meantime, have a great day my friend!

[Stephen1966]

Today, I went through the PSU and changed the resistors in the dropping string. I based this on the PSU of the #060. The gain structure of the amp from preamp, through the PI to the power amp now seems more balanced. I'm posting the schematic with all the latest updates at head of the topic on the first page. The reverb is more tasteful now (a lot less like the public toilet it sounded like before) though I still get a fair bit of noise on the return pot when I turn it up. The rest of the circuit is dead quiet. The noise on the return pot could be because of lead dress - it crosses paths with the PSU supply to the OD and Clean channels and it has something like a 60Hz hum to it. Though it could be because it is sitting close to the variac. The #060 had a 1k8 dropping resistor between the clean and OD nodes but I had a voltage difference of just around three volts with that, so I increased it to 3k6 and now there is about ten volts difference, which I believe is more in the ball park we are looking for. V1 is a more balanced tube whereas V2 has a five volt difference between the triodes. There's plenty of grainy structure to the sound there so I'm not unhappy with it. I also swapped out the 100k plate resistors for [Dale] RN65Ds - can't say I noticed much of a difference there but it's not unpleasant. The tone is what I would call more rounded but it probably needs rebiasing and the PI adjusting some more. I have some more parts on the way, as well... played it all kinds of ways today but I'm still looking for the bloom. With the gain up in OD with the PAB engaged, there is the suggestion of bloom but nothing remarkable yet. Tomorrow, I will be replacing the master and volume pots with j-taper pots. I don't imagine that is going to do much to the sound but it should improve the response of the volume and master controls.

Some further thoughts...

I was expecting a step up in tone from the Dale plate resistors but it wasn't a huge difference from the Vishay/Beyschlags I had been using (Mouser: MBE04140C1003FC100). The effect is a sweeter, more rounded tone perhaps, all of which is difficult to explain and even harder to quantify. The Dales are all good high tolerance resistors and readily available - measuring 99.9k generally, putting them within 0.1% tolerance. I think here, with the Vishay/Beyschlags we probably have a very similar, equivalent sounding resistor. That said, the Dales are readily available and are said to make a great contribution to the signature tone of the amp so there is no pressing need to use an equivalent - I should have just used them from the beginning. Others might say differently. I'm always aware that what I hear, may not be what others can hear. It's part of the reason why I think it would be futile to offer a BOM for this project. If you get into building this project, it's assumed you will already have your own ideas about which parts sound the best. This was just an interesting comparison.

[Stephen1966]

The j-taper pots are in - very nice. Thank you for the suggestion. They allow a much more sensible sense of control. A good level of volume with master and volume at less than 50% on the dial.

[Stephen1966]

I'm back but only briefly... pressures of work are keeping me from the amp. I did get hold of a new oscilloscope with FFT though and I'm getting acquainted with it now. I'd like to see if I can track down the second harmonic but as the FFT only operates on one of the scope's channels I was wondering if there was a procedure for hooking it up to the output of the amp. I was able to get a balanced signal at the cathodes of the power tubes (described earlier) but because of the one channel limitation, I don't yet see a way of taking the phased signal from there. Is it possible I could map the frequencies from just one side of the output transformer via the cathodes?