EMI RS124
- Thread starter ed rees
- Start date
Help Support GroupDIY Audio Forum:
MagnetoSound
Well-known member
It adds up to 2.6A, so you should see about 6V okay.
I think that 1/2 an amp is way high for the B+. that's 90VA (180v 1/2A).
the output stage draws 10.5mA for both tridodes, and the input tube, plus the voltage divider that makes the threshold voltage draws 6.2mA. That's about 17mA total, so call it 25mA for old time's sake. That's a per-channel number.
the output stage draws 10.5mA for both tridodes, and the input tube, plus the voltage divider that makes the threshold voltage draws 6.2mA. That's about 17mA total, so call it 25mA for old time's sake. That's a per-channel number.
![Electronics Soldering Iron Kit, [Upgraded] Soldering Iron 110V 90W LCD Digital Portable Soldering Kit 180-480℃(356-896℉), Welding Tool with ON/OFF Switch, Auto-sleep, Thermostatic Design](https://m.media-amazon.com/images/I/41gRDnlyfJS._SL500_.jpg)
i should point out that all of that was inferred from the voltages and resistances shown on the Altec schematic. There's lots of clues to what's going on that are printed there in broad daylight. You just have to know how to interpret them.
The current through the output stage is identified by the voltage on the cathode resistor, R7 (looking at 436C schem). 2.3v across 220R is .015A
The current for V1 and the threshold divider is given up from seeing that the voltage at the junction of R3 and R4 is 210v at rest (worst case) and on the other side of R13, it's 272v, so that's 62V across 10k, which is .0062A, or 6.2mA. oh... looks like I goofed on addition, it's 21.2mA total, but rounding up to 25mA gives 4mA margin. I think I'd go up to 30-40mA in terms of the power transformer... But at least you now have a range to look thru.
It might be easier to change the transformer (meaning easier to find one that will work) by going to full-wave center tap or even a bridge rectifier.
Since you need 272v at the input to the filter system, a FW center-tap transformer would be about 193-0-193, and for a bridge, it would just be 193v. Of course, you're likely not to find that exact voltage, so you can go up or down slightly from that, and you'll have to adjust R13 and R14 to get the voltages back. I'd say that going down would be better, so you adjust R13 and R14 downward, and you'll want to make C8B and C8C larger anyway because we're not trying to squeeze every last penny out of the BOM. Lowering those values is better because it improves the regulation of the supply.
Now, someone will probably say something about part of the charm of this thing being how the operating points of the tubes change with compression, and that might be so, in which case you're back to finding the same transformer that Altec used... BTW, my schematic says 117v for the secondary, which with a doubler gets you easily to 272v. They used a selenium rectifier, so it has more voltage drop than the silicon diodes that we'd use today. If you go the 117v route, then a simple isolation transformer gets you there. I'd try to find something with double bobbin construction, so you have more isolation from the power line than if the windings were simply laid on top of each other.
The current through the output stage is identified by the voltage on the cathode resistor, R7 (looking at 436C schem). 2.3v across 220R is .015A
The current for V1 and the threshold divider is given up from seeing that the voltage at the junction of R3 and R4 is 210v at rest (worst case) and on the other side of R13, it's 272v, so that's 62V across 10k, which is .0062A, or 6.2mA. oh... looks like I goofed on addition, it's 21.2mA total, but rounding up to 25mA gives 4mA margin. I think I'd go up to 30-40mA in terms of the power transformer... But at least you now have a range to look thru.
It might be easier to change the transformer (meaning easier to find one that will work) by going to full-wave center tap or even a bridge rectifier.
Since you need 272v at the input to the filter system, a FW center-tap transformer would be about 193-0-193, and for a bridge, it would just be 193v. Of course, you're likely not to find that exact voltage, so you can go up or down slightly from that, and you'll have to adjust R13 and R14 to get the voltages back. I'd say that going down would be better, so you adjust R13 and R14 downward, and you'll want to make C8B and C8C larger anyway because we're not trying to squeeze every last penny out of the BOM. Lowering those values is better because it improves the regulation of the supply.
Now, someone will probably say something about part of the charm of this thing being how the operating points of the tubes change with compression, and that might be so, in which case you're back to finding the same transformer that Altec used... BTW, my schematic says 117v for the secondary, which with a doubler gets you easily to 272v. They used a selenium rectifier, so it has more voltage drop than the silicon diodes that we'd use today. If you go the 117v route, then a simple isolation transformer gets you there. I'd try to find something with double bobbin construction, so you have more isolation from the power line than if the windings were simply laid on top of each other.
While we have the subject of PT on the plate I'll ask the question I haven't seen asked yet, just for reference, context, and discussion.
Q: Why did Altec choose to use the lower HT PT with voltage doubler circuit instead of a standard CT with full wave rectifier?
More design related or bean counter decision?
FWIW, on my clone I used a standard full wave (230-0-230) I think, with tube rectifier and corresponding resistance changes in PS as needed. Works great.
Also, don't be surprised if your cathode voltage isn't spot on to the schematic. Mine never was through several tube subs and wound up staying at exactly the amount DaveP's did on his clone in another thread.
Q: Why did Altec choose to use the lower HT PT with voltage doubler circuit instead of a standard CT with full wave rectifier?
More design related or bean counter decision?
FWIW, on my clone I used a standard full wave (230-0-230) I think, with tube rectifier and corresponding resistance changes in PS as needed. Works great.
Also, don't be surprised if your cathode voltage isn't spot on to the schematic. Mine never was through several tube subs and wound up staying at exactly the amount DaveP's did on his clone in another thread.
my guess is that it was bean counter related, and it probably made using the selenium rectifier easier... They did a similar supply in the 1567 too.
it was a novel circuit, and popular at the time.
yes, the cathode voltage is related to the tube's transconductance, and that is a function of manufacturing and tube lifespan. The voltages are never intended as go/nogo, rather they indicate about what you can expect there. If circuits depended on voltages being exact, things would be much harder to manufacture. It's the goal in engineering to produce a product whose performance is nearly independent of variations in tubes/transistors/etc. Components have to be within their tolerance, but within that, stuff is supposed to work.
there are times when voltages need to be exact, such as the reference voltage of a DAC, but those should be the exception, not the rule.
it was a novel circuit, and popular at the time.
yes, the cathode voltage is related to the tube's transconductance, and that is a function of manufacturing and tube lifespan. The voltages are never intended as go/nogo, rather they indicate about what you can expect there. If circuits depended on voltages being exact, things would be much harder to manufacture. It's the goal in engineering to produce a product whose performance is nearly independent of variations in tubes/transistors/etc. Components have to be within their tolerance, but within that, stuff is supposed to work.
there are times when voltages need to be exact, such as the reference voltage of a DAC, but those should be the exception, not the rule.
Assuming the bean counter idea, it makes things like this kind of ironic.
http://www.guitarpartssite.com/Mercury-Magnetics-ALTEC-436-PT-MM-p/altec-436-pt-mm.htm
Though I know that some are interested in strict restoration to exact spec.
I wanted to make the roundabout point to DIY newcomers that you can build a successful 436 unit with any number of off the shelf PTs and not miss the mark on why people like these. And, as rickc pointed out, tube circuits will pretty much always show a small variation in circuit voltages that are of no major concern.
http://www.guitarpartssite.com/Mercury-Magnetics-ALTEC-436-PT-MM-p/altec-436-pt-mm.htm
Though I know that some are interested in strict restoration to exact spec.
I wanted to make the roundabout point to DIY newcomers that you can build a successful 436 unit with any number of off the shelf PTs and not miss the mark on why people like these. And, as rickc pointed out, tube circuits will pretty much always show a small variation in circuit voltages that are of no major concern.
I have tried to modify my altec clone, but ... I think that the best thing is to built a new one!
There is someone that has designed only the RS124 output stepped attenuator?
Really I don't understand it: I am not a tech and a separate schematic of the attenuator with input and output terminals may be really useful
There is someone that has designed only the RS124 output stepped attenuator?
Really I don't understand it: I am not a tech and a separate schematic of the attenuator with input and output terminals may be really useful
ok. I looked at what was on Rotheu's schematic and then did my own bridged-tee attenuator to run at 300R. The 436 delivers a lot of output, and unless you're following it with something lossy, it's probably better to take a 6dB level hit by strapping it for 300R, and then working from there.
I will post a schematic as soon as I can get it drawn up.
I will post a schematic as soon as I can get it drawn up.
here is the output pad (mine) for the RS124. I don't know what was done before (and I don't care).
Caveats:
The rotary switch is ganged (one shaft, 2 switches), and ideally it should be make before break, but these can be difficult to get. Using a break-before-make switch here means it will not be silent if you operate it with audio passing thru. It won't go BANG at 180dB, but it will make a soft click. This just means that you shouldn't change the setting when the music is running.
Resistor values are standard e-96/1% values. Resistors can be 1/4W or 1/2W metal film.
Caveats:
The rotary switch is ganged (one shaft, 2 switches), and ideally it should be make before break, but these can be difficult to get. Using a break-before-make switch here means it will not be silent if you operate it with audio passing thru. It won't go BANG at 180dB, but it will make a soft click. This just means that you shouldn't change the setting when the music is running.
Resistor values are standard e-96/1% values. Resistors can be 1/4W or 1/2W metal film.
that's the schematic posted by "rotheu"
I didn't try to analyze what was done there. I took his 300R output impedance ( and now that I'm awake, I think Zo is 150R with the secondaries paralleled). confirmed.
Rotheu's values look like it's a non-constant impedance design, and what I did was to do a constant impedance design, with his attenuation values, and 300R impedance. That means the output transformer and output stage won't be loaded quite as heavily, but still are loaded. if it really concerns you, then you could add 300R in parallel with the output of the 436, which with the attenuator's 300R input z, gives a 150R load on the unit.
My attenuation values are the same that EMI chose, 0, -5, -10, -15, -20, -30dB. The attenuator is textbook, with values converted to nearest 1% standard values.
I didn't try to analyze what was done there. I took his 300R output impedance ( and now that I'm awake, I think Zo is 150R with the secondaries paralleled). confirmed.
Rotheu's values look like it's a non-constant impedance design, and what I did was to do a constant impedance design, with his attenuation values, and 300R impedance. That means the output transformer and output stage won't be loaded quite as heavily, but still are loaded. if it really concerns you, then you could add 300R in parallel with the output of the 436, which with the attenuator's 300R input z, gives a 150R load on the unit.
My attenuation values are the same that EMI chose, 0, -5, -10, -15, -20, -30dB. The attenuator is textbook, with values converted to nearest 1% standard values.
ok. as near as I can tell, the two schematics are the same.
I'm not too impressed with the output attenuator in the design.
I think that this counts on having a 600R load afterwards, otherwise the steps are even further off.
the steps aren't terribly accurate (and I'm not nit picking over fractions of a dB)
the -30 step is actually OFF
the impedance presented to the 436 varies wildly depending on attenuation (772R to 168R)
I did a spreadsheet for this, so I could play with some assumptions.
I'm going to forget that I ever saw this part of the diagram.
I'm not too impressed with the output attenuator in the design.
I think that this counts on having a 600R load afterwards, otherwise the steps are even further off.
the steps aren't terribly accurate (and I'm not nit picking over fractions of a dB)
the -30 step is actually OFF
the impedance presented to the 436 varies wildly depending on attenuation (772R to 168R)
I did a spreadsheet for this, so I could play with some assumptions.
I'm going to forget that I ever saw this part of the diagram.
