how about an "extended class A" approach?
2 PP class A triodes in parallel with a pair of pentodes class AB. The triodes do all the low power work and the pentodes only kick-in for higher powers.
I don't know a lot about this topology, but it interests me. Maybe someone else here can comment on the pros/cons of this output stage. I know its briefly mentioned in RDH,4ed.
2 PP class A triodes in parallel with a pair of pentodes class AB. The triodes do all the low power work and the pentodes only kick-in for higher powers.
I don't know a lot about this topology, but it interests me. Maybe someone else here can comment on the pros/cons of this output stage. I know its briefly mentioned in RDH,4ed.
> the pencil and paper idea fills me with trepidation
ALL the Classic Era amplifiers were designed with pencil, paper, and slide-rule.
You can use a digital calculator, but it leads to lazy thinking. And it gets you too close to the VLSI demon that made SPICE both possible and necessary.
The implication of such design is that you don't jerk-around designing. You rough-out the plan and then go to the breadboard. True breadboarding is another lost art. You need a construction system solid enough and tight enough to give good results, yet flexible enough that you can quickly rip-out large sections and try something else.
ALL the Classic Era amplifiers were designed with pencil, paper, and slide-rule.
You can use a digital calculator, but it leads to lazy thinking. And it gets you too close to the VLSI demon that made SPICE both possible and necessary.
The implication of such design is that you don't jerk-around designing. You rough-out the plan and then go to the breadboard. True breadboarding is another lost art. You need a construction system solid enough and tight enough to give good results, yet flexible enough that you can quickly rip-out large sections and try something else.
I was being facetious about the "pencil and paper" suggestion. I agree that there's no substitute for building the thing and finding out what actually works. However, I don't have an extensive parts bin for trial-and-error experiments and there is a lack of access to reliable parts suppliers in this country. Thus, I'm not in the happy position to be able just to try lots of alternative designs and measure them to see what works. For this reason, I find this forum extremely helpful for seeking advice from people who've "been there, done that".
Design and planning has to come first, naturally, using the most appropriate tools available at the time. I designed my current amp before I ever got near Spice, using an Excel spreadsheet. I drew the schematic on one worksheet (pure Microsoft graphics, with no intelligence) and did the calculations for voltages, currents, gain, bandwidth, feedback etc, on a second worksheet. This, to me, is a more modern equivalent of pencil, paper and slide rule (which were my tools in the 1960s). I didn't consider a slide rule was cheating then and I don't consider a spreadsheet to be cheating now, either. It only does what you tell it, after all, so it doesn't take away the thinking part, just makes it more convenient when you want to make changes to the design.
I only got my hands on LTspice a month or so ago and found it to be great fun. However, I don't really know how far one can go with Spice modeling, in terms of getting believable, real-world results. It obviously depends on how faithfully the models follow the behaviour of the tubes concenred and it can't, of course, show the effects of individual tube variations from standard specs, hum, RF pickup and other peculiarities of individual components and layout.
Design and planning has to come first, naturally, using the most appropriate tools available at the time. I designed my current amp before I ever got near Spice, using an Excel spreadsheet. I drew the schematic on one worksheet (pure Microsoft graphics, with no intelligence) and did the calculations for voltages, currents, gain, bandwidth, feedback etc, on a second worksheet. This, to me, is a more modern equivalent of pencil, paper and slide rule (which were my tools in the 1960s). I didn't consider a slide rule was cheating then and I don't consider a spreadsheet to be cheating now, either. It only does what you tell it, after all, so it doesn't take away the thinking part, just makes it more convenient when you want to make changes to the design.
I only got my hands on LTspice a month or so ago and found it to be great fun. However, I don't really know how far one can go with Spice modeling, in terms of getting believable, real-world results. It obviously depends on how faithfully the models follow the behaviour of the tubes concenred and it can't, of course, show the effects of individual tube variations from standard specs, hum, RF pickup and other peculiarities of individual components and layout.
Good rebuttal.
And you do know slide-rules. You learned to think.
I still feel that owning a dozen calculators has rotted my math "feeling", so I do keep a sliderule handy. It is good to be reminded that all the wrong answers are available, and just because the cursor or LED is showing a number does not mean it is right. (And the slipstick is a killer tool for ratios.)
Sliderules also remind us that 8-place precision is pointless... just the other day I was looking at a plan that showed "7.15mA" and it took a while to sink in that it will never be 7.15mA. Working it my own way, with maybe too much tube experience, I got "7mA to 9mA", and I would not be shocked to find a tube that gave 5mA in that circuit.
I overlooked your location. FWIW, in the USA near a large city, I find it easiest to order tubes from another country (Canada). It's a global world. But I grant that the US/Canada border is more open than many. In fact the main delay isn't getting stuff from Canada or California to near the NYC area, it is that if you don't PAY for overnite service, UPS and FedEx hold the box in Newark until the scheduled delivery-time for the service you paid for.
And you do know slide-rules. You learned to think.
I still feel that owning a dozen calculators has rotted my math "feeling", so I do keep a sliderule handy. It is good to be reminded that all the wrong answers are available, and just because the cursor or LED is showing a number does not mean it is right. (And the slipstick is a killer tool for ratios.)
Sliderules also remind us that 8-place precision is pointless... just the other day I was looking at a plan that showed "7.15mA" and it took a while to sink in that it will never be 7.15mA. Working it my own way, with maybe too much tube experience, I got "7mA to 9mA", and I would not be shocked to find a tube that gave 5mA in that circuit.
I overlooked your location. FWIW, in the USA near a large city, I find it easiest to order tubes from another country (Canada). It's a global world. But I grant that the US/Canada border is more open than many. In fact the main delay isn't getting stuff from Canada or California to near the NYC area, it is that if you don't PAY for overnite service, UPS and FedEx hold the box in Newark until the scheduled delivery-time for the service you paid for.
Alas, ordering from overseas is fraught with problems here.
Firstly, many Internet suppliers don't accept orders from Indonesia, because of the copious credit card fraud. I can't blame them for that.
Those suppliers who do, such as Amplimo and Antique Electronic Supply, tend to use the normal postal service unless specifically requested to use a courier. Using the postal service is asking for trouble here. Things can disappear or be "held to ransom" (usually stolen from the customs area) so that you have to pay several times the correct import duty to retrieve them.
Another little trick of the Indonesian post office is to return to sender, automatically, any items from overseas that are not collected within two weeks. However, they usually notify the addressee only one day (or even less!) before the collection deadline. They will not deliver such items and the addressee always has to pay a surcharge, regardless of how much postage was paid by the sender. UPS, DHL etc. are safe and quick but expensive; for a $100 order, it doesn't make much sense to use their services.
Having said all that, I have successfully received three consignments each from both Antique and Amplimo during the past year or so, only one of which was held to ransom. Most of them were notified almost too late to be collected, however.
Firstly, many Internet suppliers don't accept orders from Indonesia, because of the copious credit card fraud. I can't blame them for that.
Those suppliers who do, such as Amplimo and Antique Electronic Supply, tend to use the normal postal service unless specifically requested to use a courier. Using the postal service is asking for trouble here. Things can disappear or be "held to ransom" (usually stolen from the customs area) so that you have to pay several times the correct import duty to retrieve them.
Another little trick of the Indonesian post office is to return to sender, automatically, any items from overseas that are not collected within two weeks. However, they usually notify the addressee only one day (or even less!) before the collection deadline. They will not deliver such items and the addressee always has to pay a surcharge, regardless of how much postage was paid by the sender. UPS, DHL etc. are safe and quick but expensive; for a $100 order, it doesn't make much sense to use their services.
Having said all that, I have successfully received three consignments each from both Antique and Amplimo during the past year or so, only one of which was held to ransom. Most of them were notified almost too late to be collected, however.
I can't seem to find what I am looking for by using the search.
I've had vacuum tube experience, but not since about 1970. However, as I understand it, the principals of physics have not changed very much since then, so that should be no problem. However, I've never used breadboards.
I'd like info on using breadboards, where to get them, etc.
And anyone provide that info, or refer me to a thread that has that info?
Thanx.
I've had vacuum tube experience, but not since about 1970. However, as I understand it, the principals of physics have not changed very much since then, so that should be no problem. However, I've never used breadboards.
I'd like info on using breadboards, where to get them, etc.
And anyone provide that info, or refer me to a thread that has that info?
Thanx.
Max Robinson, at his "Fun with Tubes" website, has 2 articles on breadboarding: 4 methods of breadboarding and
Best breadboard.
Hope that helps.
Best breadboard.
Hope that helps.
Revisiting this old thread, FWIW I thought I might say a bit more about the EL34 triode Class AB2 idea I was asking about.
I can think of three advantages to be gained by providing a direct-coupled cf or sf driver for PP EL34s in triode, UL or pentode connection, as follows:
1. Avoid any risk of blocking distortion that might otherwise occur on transients with capacitive coupling to the OP tubes.
2. Provide a high impedance load to the preceding stage, which could be a splitter or a PP or differential stage, thus enabling a relatively whimpy but high gain tube to be used, such as 6SL7 or 12AX7.
3. Enable the OP stage to stray into AB2 on transients without drama. I don't believe that EL34 control grids would be robust enough to tolerate sustained grid current, but my guess would be that an occasional excursion into A2 on transients wouldn't hurt.
I can think of three advantages to be gained by providing a direct-coupled cf or sf driver for PP EL34s in triode, UL or pentode connection, as follows:
1. Avoid any risk of blocking distortion that might otherwise occur on transients with capacitive coupling to the OP tubes.
2. Provide a high impedance load to the preceding stage, which could be a splitter or a PP or differential stage, thus enabling a relatively whimpy but high gain tube to be used, such as 6SL7 or 12AX7.
3. Enable the OP stage to stray into AB2 on transients without drama. I don't believe that EL34 control grids would be robust enough to tolerate sustained grid current, but my guess would be that an occasional excursion into A2 on transients wouldn't hurt.
I hear you.
I'd like to try 6072 driving 300B using something like this.
It would be like a gyrator to the driver anode, a SF driving
the power tube grid, and a source of fixed bias for the power
tube.
Michael
PS The grid drive supply just needs to be 30-40 volts returned to
the cathode of the power tube to keep the grid current loop local.
Or it can be derived off B+ depending on the cathode circuit tolerance
for extra current...
I'd like to try 6072 driving 300B using something like this.
It would be like a gyrator to the driver anode, a SF driving
the power tube grid, and a source of fixed bias for the power
tube.
Michael
PS The grid drive supply just needs to be 30-40 volts returned to
the cathode of the power tube to keep the grid current loop local.
Or it can be derived off B+ depending on the cathode circuit tolerance
for extra current...
Attachments
ray_moth said:
Thank you, Ray; I bookmarked the links.
I was hoping to find some system that does not require soldering since that would make it easier to make changes. However, if solderless breadboards worked well, then the links surely would have recommended them.
Regarding making a secure mechanical connection before soldering, I don't except in situations where there could be some mechanical strain on the connection. Too many times I've had the problem of un-soldering connections in which someone has wrapped the wire around a terminal several times. If one has to un-solder something that is heat sensitive, it can easily be destroyed by the time one gets it un-soldered if the wire has been wrapped. In my opinion, if soldering is done properly, it will hold without depending on wrapping wires.
- Status
- Not open for further replies.