In my PP amp build I'm finding that I'm running out of space (moving bits around chassis, not constructing yet).
Amp has separate power trannies for each B+, plus a bias supply tranny, output tube filament trannies for each channel and a shared filament tranny for the input tubes. (overkill, but they are the trannies I had access to)
This means stuffing 8 trannies onto or into the chassis before I even start with caps, bias pots, sockets etc.
As the front end of the amp is just an input selector and a switched attenuator, I'm thinking that there's barely any difference between what I'm building and a passive pre with two monoblocks.
It'd certainly make for a far neater layout (and I have a couple of spare 2 x 12v AC toroids that I can use to feed both the input tubes and each channel's neg bias supply).
Is there a good reason not to go this way?
Amp has separate power trannies for each B+, plus a bias supply tranny, output tube filament trannies for each channel and a shared filament tranny for the input tubes. (overkill, but they are the trannies I had access to)
This means stuffing 8 trannies onto or into the chassis before I even start with caps, bias pots, sockets etc.
As the front end of the amp is just an input selector and a switched attenuator, I'm thinking that there's barely any difference between what I'm building and a passive pre with two monoblocks.
It'd certainly make for a far neater layout (and I have a couple of spare 2 x 12v AC toroids that I can use to feed both the input tubes and each channel's neg bias supply).
Is there a good reason not to go this way?
three times the chassis to make!
Upside - you get to fudge up a whole lot of sexy looking handmade interconnects.
Good to hear its all progressing Drew...
Upside - you get to fudge up a whole lot of sexy looking handmade interconnects.
Good to hear its all progressing Drew...
Sounds like you have a couple of options.
1. Reduce number of transformers.
2. increase number of chassis.
Why so many transformers?
You may want to consider as an option a separate power supply chassis, and then a single amp chassis with the passive switching and volume control.
1. Reduce number of transformers.
2. increase number of chassis.
Why so many transformers?
You may want to consider as an option a separate power supply chassis, and then a single amp chassis with the passive switching and volume control.
I like Chris' suggestion, as it deals with the interconnect cable issue associated with passive control centers. By definition, a passive control center has a high O/P impedance and poor drive capability. That forces the use of short, low capacitance interconnect cables between the control center and power amp. Mounting the control center in the power amp chassis, an "integrated amp", bypasses the interconnect cable problem.
I had pondered Chris's option of a separate PSU too, but with an under 5 in the house, the idea of a HT umbilical concerns me. Far safer to "umbilical" a line level signal. Will not be a long cable run.
As for why so many transformers, it's simply what I had available to me. To be sure, a single mains with B+, a neg bias supply and enough 6.3v AC to run everything would take way less space, but twin B+ trannies without heater or bias is what I've got here.
As for why so many transformers, it's simply what I had available to me. To be sure, a single mains with B+, a neg bias supply and enough 6.3v AC to run everything would take way less space, but twin B+ trannies without heater or bias is what I've got here.
Regarding HT umbilical, in my opinion, it is probably safer than any valves on display, that is, if the valves are not completely enclosed in the chassis. A child could hit the valve, break the glass and be exposed the the B+ on the plates. If you use a HT umbilical, check out Building Valve Amplifiers by Morgan Jones if you have it, he addresses some of the safety issues. For instance, make sure that the HT end of any connector is the female component so that no fingers can touch HT if disconnected.
Not sure how much space it saves, but the bias can be tapped from the B+ transformer, saving at least that transformer. See how they do it on guitar amps http://www.ax84.com/static/corepoweramps/20W_PP/AX84_20W_PP_Poweramp_Schematic.pdf
Maybe if you post your intended design/schematic and the specifications of your available components someone here might be able to offer some other advice.
Good luck with the project!
Chris
Not sure how much space it saves, but the bias can be tapped from the B+ transformer, saving at least that transformer. See how they do it on guitar amps http://www.ax84.com/static/corepoweramps/20W_PP/AX84_20W_PP_Poweramp_Schematic.pdf
Maybe if you post your intended design/schematic and the specifications of your available components someone here might be able to offer some other advice.
Good luck with the project!
Chris
once correct balance is found it stays there 😉 but hey, isnt it about very enthusiastic amps 😀
Short cabling between control center and monoblocks is not enough. Shielded wiring, with its comparatively high pF./ft. is out. Braided, unshielded, cabling is necessary, to hold the capacitance down. A 3 wire braid, with 2 "cold" conductors and 1 "hot" conductor, should be OK.
Last edited:
What you are doing is trading bulk for cost. If you use the existing transformers you end up with an amp that is physically huge and in multiple chassis but you savethe cost of a new transformer. But what do you pay for the extra chassis? You really you only save the difference between the cost of a chassis and transformers. What is that $50?
The B+ power cord can be safe. What you do is place al the chassis into one cabinet. It can be compact too. Place one chasses on the bottom, tubes pointed up and place the other chasis, the one with the controls hung from the top, tubes, transformers and whatever pointed down. (use tube retainers).
If an open top is required mount the chassis vertically with the tubes in horizontal possition. most tubes are rated for "any position". There are many combination. A wood cabinet has 6 sides and you can mount chassis on any of them. This way the high voltage wire and the hot tubes are contained. So they are safe for kids and cats.
I'm building an amp now and it uses a cabinet that has wooden rails inside. Standard Hammond 16x8x2 chassis slides in between rails made on 3/4" square wood and is held in place by a screw
I've dismantled a few old Hammond organs. They are all very well designed for service and have multiple steel chassis inside a wood cabinet. many tinme the chassis will be mounted to the wood in a hinge so it can swing out to expose the guts
The typical HiFi DIY amp is just a bare chassis. that works only if it is just a one chassis design. For two or more you need some woodwork
One more thing. A passive preamp (if that is not an oxymoron) compleely disapears if yu have a wood cabinet. It becomes just some controls mounted on a front panel that with shielded wire going back the main amp chassis(s).
The B+ power cord can be safe. What you do is place al the chassis into one cabinet. It can be compact too. Place one chasses on the bottom, tubes pointed up and place the other chasis, the one with the controls hung from the top, tubes, transformers and whatever pointed down. (use tube retainers).
If an open top is required mount the chassis vertically with the tubes in horizontal possition. most tubes are rated for "any position". There are many combination. A wood cabinet has 6 sides and you can mount chassis on any of them. This way the high voltage wire and the hot tubes are contained. So they are safe for kids and cats.
I'm building an amp now and it uses a cabinet that has wooden rails inside. Standard Hammond 16x8x2 chassis slides in between rails made on 3/4" square wood and is held in place by a screw
I've dismantled a few old Hammond organs. They are all very well designed for service and have multiple steel chassis inside a wood cabinet. many tinme the chassis will be mounted to the wood in a hinge so it can swing out to expose the guts
The typical HiFi DIY amp is just a bare chassis. that works only if it is just a one chassis design. For two or more you need some woodwork
One more thing. A passive preamp (if that is not an oxymoron) compleely disapears if yu have a wood cabinet. It becomes just some controls mounted on a front panel that with shielded wire going back the main amp chassis(s).
That equation may hold for you in the USA Chris, but down in this hemisphere we have a completely different economic reality. Iron is generally hard to get in usable formats, and one-off importing eg Hammond or Edcor and the like, is prohibitively expensive - the price of a single transformer can treble once freight is added.
Chassis stuff we have plenty of!
This one in particular strikes me as an interesting comment. The source selector and volume attenuator in my amp come before the first 6 6CG7 triode gain stage. effectively they are a passive preamp, just connected by wires to the inputs of the power amp, rather than exiting through rca sockets.
As has already been stated, this significantly reduces the capacitance that the signal sees.
Without access to drawing software, I'll have to explain the cct configuration but it's hardly earth shattering:
Input selector goes to 100k switched attenuator. Output of attenuator has a 1M between hot and ground rail. Signal from hot side enters the grid of a 6CG7 via a 2k2 grid stopper. B+ is 300v, anode resistor is 47k, cathode is biased with 3 LED's to ground.
Output from anode of 6CG7 goes through a 0.33uf cap to a second half of the 6CG7. This is configured as a concertina splitter with LED bias. 300v B+, 22k above anode, 3 x LED's and 22k to ground from cathode. bias voltage from below the LED's taken to grid via a 470k resistor.
Outputs from the two halves of the concertina feed the grids of triode connected EL84's via 0.33uf caps. B+320v, 10K primary PP tranny, fixed bias.
No loop feedback applied.
Pretty much the minimal cct required to make things function, no cathode bypass caps anywhere.
What I mean by "completely disappear" is that while the parts and functions remain there is no longer a thing that lives in a chassis with RCA connectors on the back that you can point to and say "This is a passive preamp". Also the space that it takes and the RCA jack are gone too and also many of it's disadvantages too, long those interconnects are gone as well.
I chose the word carefully. "Disappear" does not mean to not exist. My car keys and money in my wallet "disappears" from time to time. It means "is gone from sight".
My main point was that multiple chassis are a good practical design idea but only if you are willing to build a cabinet to house them. the cabinet turns something that looks like a lab experiment into a product.
Gotcha!
Yes, happy to build a cabinet, as Ash has alluded, wood and construction time is easy to get hold of here, getting good iron is more of a pain in the bott.
Comments on the proposed schema as described are welcome.
This is generally how I do it for my own use, and results in extremely quiet (electrically speaking) amplifiers.
As you get older placing the heavy components on multiple chassis will allow you to move things without hurting yourself - even if it does take a bit more space. I'm starting to be glad that I did this, and I have friends who built huge amps on single chassis and no longer have the strength to move them for various reasons.
I use a passive pre (selector switch + 10k pot in a box) with a 3ft RCA's on the output feeding mono block amps. when using a typical low impedance source, the largest source impedance the RCA cable will see is aprox 5k (when the pot turned to center resistance position).
Twisted pair cable would work better than shielded, but the worst shielded 3ft RCA cable I've measured was 100pF. Using the worst offending cable I have gives a -3dB point at 318kHz when the pot is in it's highest Z part of the rotation if you don't account for the amps Cin. Something to be mindful of, but I wouldn't call shielded cable completely off limits, unless the amp has a large Cin. I also wouldn't use shielded cable if your forced to use an attenuator larger in value than 10-20k ish.
For a DIY'er a separate active or passive pre is much more flexible long term than going the integrated route. A separate pre can be used with future projects whether they be power amps, mono-blocks or prototype bread board amps. You won't have to wire a new selector switch + attenuator for each project. And if you do ever build mono blocks (for this project or the next) have fun adjusting the volume using separate built in attenuators.
Twisted pair cable would work better than shielded, but the worst shielded 3ft RCA cable I've measured was 100pF. Using the worst offending cable I have gives a -3dB point at 318kHz when the pot is in it's highest Z part of the rotation if you don't account for the amps Cin. Something to be mindful of, but I wouldn't call shielded cable completely off limits, unless the amp has a large Cin. I also wouldn't use shielded cable if your forced to use an attenuator larger in value than 10-20k ish.
For a DIY'er a separate active or passive pre is much more flexible long term than going the integrated route. A separate pre can be used with future projects whether they be power amps, mono-blocks or prototype bread board amps. You won't have to wire a new selector switch + attenuator for each project. And if you do ever build mono blocks (for this project or the next) have fun adjusting the volume using separate built in attenuators.
Last edited:
- Status
- Not open for further replies.