Feeler: Classic Solo

[ruffrecords]

Well done. Very neat build. Is it OK if I pass the pics to Pierre who keeps an independent gallery of builds based on my designs?

Cheers

Ian

 

[3nity]

Yes off course Ian.

 

[NorthernLight]

Hi Ian,

I am currently collecting information for a future two channel Pultec build. I have all the info I need about the passive EQ section, and am now venturing into tube gain makeup territory. This would be my first tube project. I do have some experience with other DIY builds, including modifying them, but in this case I am a bit out of my depth. I hope you don't mind if I ask some questions that might have been answered before - I have read pages upon pages in both the Pultec and Classic Solo threads and have not found the definitive answers I am looking for.

Apparently it is possible to use a Classic Solo PCB as a PSU and Gain Make Up solution for a two channel Pultec EQ build (instead of two PMTGMU boards and a separate PSU board). I am a fan of less PCBs to mount, so this sounds apealing. I have examined the Classic Solo PCB layout and schematic in the documentation and my intuition says I need to look at the Dual DI setup to achieve this.

For reference: I am looking to use Carnhill VTB9071 10K/10K input transformers and Carnhill VTB2291 600/600 output transformers, as derived from the Pultec documentation.

The Classic Solo documentation describes a modification of using the 6922 in an SRPP configuration instead of a MU follower configuration. From what I read, SRPP has lower gain than MU follower and SRPP produces more distortion before it starts clipping, whereas MU follower has more gain by default but clips earlier. Does the lower gain of the SRPP configuration even matter if the Classic Solo is used as a Gain Make Up stage for a Pultec EQ? The behaviour of the SRPP (more distortion before clipping) sounds desirable somehow.

It would be nice to be able to dial in (extra) coloration. If I look at the schematic of the Single DI configuration, this is done by feeding the signal through the second tube and using a 10K LOG pot as a gain control (I guess you could call this "drive" control?). I think this would mean that a second Classic Solo PCB is needed for a two channel Pultec EQ (and thus twice the tubes and other components, minus one PSU section). The driven signal is attenuated by a 1K LOG potentiometer. Would the extra attenuator be needed if only on tube stage per channel is used (thus no "drive" control)?

Best regards, Alex

 

[ruffrecords]

HI Alex,

The classic solo will work fine as a gain make up amplifier for a Putlec EQ. You need at most 20dB of gain for this and both the SRPP and Mu Follower versions have more gain than this.

You are right in thinking you can obtain a lot of colouration by adding another SRPP stage and driving it hard and then attenuating the output. And it will need an extra classic solo board.

The big question is how much colour do you need. The SRPP has just enough gain to make up the EQ loss and the 6dB loss in the output transformer with a little to spare so under normal conditions it will be working at around your nominal line level (typically +4dBu). The distortion in an SRPP is directly proportional to output level and is typically 3.5% at a level of +20dBu falling to about 0.35% at 0dBu so there is already a nice bit of colouration at normal operating levels. The easiest way to add more colur is to simply overdrive the EQ input and add the 1K [ot across the SRPP's transformer secondary to restore normal operating level.

Cheers

Ian

 

[NorthernLight]

Hi Ian,

so in practice, just feed a hotter signal into the box for more colour.

The attenuator; I assume that would be a dual gang 1K LOG pot like depicted on page 13, but instead of using it as a pad on the input transformer, it would be connected in the same way on the output transformer? The schematic on page 13 also shows resistor R3 of 68 Ohm, would an additional resistor also be necessary in this application? If so, what value would be best?

Cheers, Alex

 

[ruffrecords]

Hi Ian,

so in practice, just feed a hotter signal into the box for more colour.

Yes, that is right

The attenuator; I assume that would be a dual gang 1K LOG pot like depicted on page 13, but instead of using it as a pad on the input transformer, it would be connected in the same way on the output transformer? The schematic on page 13 also shows resistor R3 of 68 Ohm, would an additional resistor also be necessary in this application? If so, what value would be best?

Cheers, Alex

I am not sure which page 13 you are referring to. Can you please clarify?

Edit: I just realised which page 13 you are referring to - the build instructions. Page 13 show a variable mic input pad whereas what you need is a variable output pad. Two versions of a suitable circuit are shown on page 23 but these days I just use a straightforward 1K LOG pot wired across the output transformer secondary.

Cheers

Ian

 

[NorthernLight]

Yes, that is right

I am not sure which page 13 you are referring to. Can you please clarify?

Edit: I just realised which page 13 you are referring to - the build instructions. Page 13 show a variable mic input pad whereas what you need is a variable output pad. Two versions of a suitable circuit are shown on page 23 but these days I just use a straightforward 1K LOG pot wired across the output transformer secondary.

Cheers

Ian

Thank you for the explanation Ian, and apologies; I could have been clearer with my question.

The output pad on page 23 of the build manual looks simple enough, it's just a dual gang LOG pot. Since this pot attenuates the balanced signal I assume this one comes after the output transformer, thus giving the option of running the entire unit hotter including both transformers? If so, it keeps getting better and better .

I haven't been able to find the PCB files for the Classic Solo Lo Hum on your site, and the PCB shows on stock even though the page says that the table was last updates in may 2021. Does this mean I can buy one from you? If so I will send you an email, as instructed on the page.

Cheers, Alex

 

[ruffrecords]

Thank you for the explanation Ian, and apologies; I could have been clearer with my question.

The output pad on page 23 of the build manual looks simple enough, it's just a dual gang LOG pot. Since this pot attenuates the balanced signal I assume this one comes after the output transformer, thus giving the option of running the entire unit hotter including both transformers? If so, it keeps getting better and better .

That is exactly what it is there for.

I haven't been able to find the PCB files for the Classic Solo Lo Hum on your site, and the PCB shows on stock even though the page says that the table was last updates in may 2021. Does this mean I can buy one from you? If so I will send you an email, as instructed on the page.

Cheers, Alex

I don't sell many PCBs these days so unfortunately I rarely update that section of the web page. As you know I have open sourced a lot of them. I am pretty sure I have a few classic solo PCBs in stock.

Cheers

Ian

 

[NorthernLight]

I don't sell many PCBs these days so unfortunately I rarely update that section of the web page. As you know I have open sourced a lot of them. I am pretty sure I have a few classic solo PCBs in stock.

I saw you open sourced many projects, much respect for that! I guess the Classic Solo is not one of them (yet). Nevertheless I'd love to support your work by buying a PCB from you. I'll send you an email to work out the details.

I also saw that there is some pretty decent build documentation for the latest revision of the Pultec EQ. When I get to this build (it is next on my ToStart list, however I have some projects on my ToFinish list so it might take a couple of months) I will make an effort to document the application of the Classic Solo PCB in SRPP configuration as a tube gain makeup for the Pultec EQ and share it here.

I have also compiled a BOM for this project with parts that are currently available from Mouser.

Regarding the BOM of small parts, I have one more question: you opted to put two 33K 2 Watt resistors in series. I guess a 66K resistor is not a common value, but the 33K part from the original Farnell BOM has a 5% tolerance, thus putting 2 in series could in practice yield a total resistance anywhere from 66.7K to 69,3K. 68K is a common resistor value and available in 2 Watt 1% tolerance, that would provide a single resistor with a value between 67,32K and 68,68K, within the tolerance of the 66K 5% solution. Is it really this simple and can a 68K 1% resistor be used at this point, or is this part more critical and am I missing something?

Cheers, Alex

 

[ruffrecords]

I saw you open sourced many projects, much respect for that! I guess the Classic Solo is not one of them (yet). Nevertheless I'd love to support your work by buying a PCB from you. I'll send you an email to work out the details.

So far the only amplifier I have open sourced is the PMTGMU. The others will be eventually but not just yet

I also saw that there is some pretty decent build documentation for the latest revision of the Pultec EQ. When I get to this build (it is next on my ToStart list, however I have some projects on my ToFinish list so it might take a couple of months) I will make an effort to document the application of the Classic Solo PCB in SRPP configuration as a tube gain makeup for the Pultec EQ and share it here.

That would be both interesting and useful.

I have also compiled a BOM for this project with parts that are currently available from Mouser.

Regarding the BOM of small parts, I have one more question: you opted to put two 33K 2 Watt resistors in series. I guess a 66K resistor is not a common value, but the 33K part from the original Farnell BOM has a 5% tolerance, thus putting 2 in series could in practice yield a total resistance anywhere from 66.7K to 69,3K. 68K is a common resistor value and available in 2 Watt 1% tolerance, that would provide a single resistor with a value between 67,32K and 68,68K, within the tolerance of the 66K 5% solution. Is it really this simple and can a 68K 1% resistor be used at this point, or is this part more critical and am I missing something?

Cheers, Alex

It is not about tolerances. I use two 33K/2W resistors in series to manage the expected heat dissipation. If the HT voltage is 300V, the total 88K string (66+22) dissipates over 1 watt of heat and three quarters of that will be in the 66K so I split that heat over two equal valued resistors. If you want to use a single 68K then I would suggest it is rated at 4W or higher.

Why not make the resistors 10 times bigger and reduce the dissipation? Two reasons. First, the few valves that specify it require a heater to cathode resistance of about 20K or less. Second the time constant of 88K with 220uF is just under 20 seconds which means even if the HT is unloaded, the output capacitor should discharge to a safe level in a reasonably short time.

Cheers

Ian

 

[NorthernLight]

It is not about tolerances. I use two 33K/2W resistors in series to manage the expected heat dissipation. If the HT voltage is 300V, the total 88K string (66+22) dissipates over 1 watt of heat and three quarters of that will be in the 66K so I split that heat over two equal valued resistors. If you want to use a single 68K then I would suggest it is rated at 4W or higher.

Hi Ian,

parts order expected to arrive early next week and can't wait to start building. I have started populating the PCB with what I had on stock, mostly the common resistors and I also ordered some tube sockets which I was able to fit on the board after bending the pins. The 6922 in the sockets should still fit (albeit snugly) in a 2U rack enclosure.

The wire bridge between P1-P3 and P2-P4 I did with wire that fit the holes on the PCB, which is 24AWG I think. Since it's coming off the HT rail I would rather have used thicker wire but I could not get the thicker wire through the hole. I then assumed is must have been designed for 24AWG but I am still hesitant; is it thick enough?

Regarding R23, I read the tread back another time and saw this question had actually already been asked and answered before. In that particular post you mention that if one chooses to use a 68K 5W resistor, it needs to be given enough spare to radiate its heat. Does that imply that putting it flat on the PCB is not a good idea?

I ordered a 5W resistor but it's slightly longer than what the lead spacing on the PCB allows for. I can either bend the pins so it is elevated a bit off the board, or I can solder the south leg of R22 through the north hole of R23 and the north leg of R23 through the south hole of R22 and put R23 flat on the PCB - those legs are connected on the bottom side of the PCB and it shouldn't become too much of a kludge that way. I think flat looks neater and I'm not a fan of exposed HT leads.

Cheers, Alex

 

[ruffrecords]

Hi Ian,

parts order expected to arrive early next week and can't wait to start building. I have started populating the PCB with what I had on stock, mostly the common resistors and I also ordered some tube sockets which I was able to fit on the board after bending the pins. The 6922 in the sockets should still fit (albeit snugly) in a 2U rack enclosure.

The PCB is designed to take 1mm gold pins for the tube sockets. I just realised the pins are listed in the BOM but not mentioned in the text.

The wire bridge between P1-P3 and P2-P4 I did with wire that fit the holes on the PCB, which is 24AWG I think. Since it's coming off the HT rail I would rather have used thicker wire but I could not get the thicker wire through the hole. I then assumed is must have been designed for 24AWG but I am still hesitant; is it thick enough?

The HT current is only about 10mA so 24 gauge wire is fine,

Regarding R23, I read the tread back another time and saw this question had actually already been asked and answered before. In that particular post you mention that if one chooses to use a 68K 5W resistor, it needs to be given enough spare to radiate its heat. Does that imply that putting it flat on the PCB is not a good idea?

Yes, flat on the PCB is not a good idea

I ordered a 5W resistor but it's slightly longer than what the lead spacing on the PCB allows for. I can either bend the pins so it is elevated a bit off the board, or I can solder the south leg of R22 through the north hole of R23 and the north leg of R23 through the south hole of R22 and put R23 flat on the PCB - those legs are connected on the bottom side of the PCB and it shouldn't become too much of a kludge that way. I think flat looks neater and I'm not a fan of exposed HT leads.

I think bending the pins so the resistor is elevated is the best solution. You could add some insulated sleeving if you are really worried about exposed HT.

Cheers

Ian