EMI RS124

[dustbro]

Could anyone tell me what the power ratings should be for R18. R19. and the 3.3M resistor in the NE2 circuit?

here's a method to determine wattage of resistors.. found it on the net, so it must be true 

The following is also known as the 'Smoke Test"
The physical size of the resistor determines how much power(wattage), or heat can be dissipated. The larger the resistor the more power it can dissipate. Resistance is the opposition to current flow(electrons) and this opposition causes heat to be generated when current flows through it.
The AMOUNT of Heat dissipated per unit of time is measured in watts.

example if P=I X E (power formula)
P=Power(watts)
E=Voltage
I=Current

ohm's law-->E=I X R

say you are using a 1/2 w 100 ohm resistor and you put 12VDC thru the resistor to ground.

determine the current: I=E/R

I=12V/100 ohms
I=120 ma.

Plug in value into ohms law formula.

P=I X E

P=.120 x 12

P=1.44 watts (resistor will be dissipating 1.44 watts at 12v with 120 milli amps flowing thru it.

Resistors come in different wattage values( 1/8w, 1/4w, 1/2w, 1w etc) usually printed on them. So for this case you would need at least a 2 WATT resistor and still be safe and NOT BURN and destroy it to a crisp.

If the resistance has not changed then your resistor passed the required safe amount of current.

If the resistance has changed you may have burned the resistor with too much current and thus you have overpowered it.

You can also use a wattmeter and measure the wattage as voltage and current is applied to the resistor at a given resistance.



Read more: http://wiki.answers.com/Q/How_do_you_determine_a_resistor's_wattage#ixzz1jOAaMYkj

 

[EarBud]

Hello Dustbro,

The "Smoke Test"... yes, I guess you can just keep increasing the wattage of the resistors until they stop blowing up  ;D.  That's pretty much where I stand now unless someone knows a less destructive way to figure it out.

As I mentioned in a previous post, I'm a good builder but not a designer.  I've always found it frustrating that so many schematics, BOM's, etc. list VALUES, but not component POWER RATINGS and things like capacitor TYPE.  I usually spend more time trying to figure these things out than any other aspect of a project.

 

[MagnetoSound]

The "Smoke Test"... yes, I guess you can just keep increasing the wattage of the resistors until they stop blowing up  ;D.  That's pretty much where I stand now unless someone knows a less destructive way to figure it out.

He just gave it to you - it's the power formula.

Measure the voltage drop across the resistor.

Use Ohm's Law to calculate the current flowing through it. I=E/R. Multiply by the voltage drop, that's the power across the resistor in Watts.


Here's another example: 22 volts across 100Ω = 0.22A.

0.22 x 22 = 4.84W.

Use 6w resistor, or better.


PS: I'm baffled - what is the NE2 circuit?

 

[Jean Clochet]

Post removed.

 

[EarBud]

Thanks everyone.

And Jean, I'm glad you brought up the neon lamp type.  I had just assumed it was an NE-2 since that's what I'm familiar with.  If you hadn't mentioned the B2A, that would have been my next question. 

 

[rickc]

I have an altec 436 that I will make RS124 modifications to. I will make a custom panel for this.

I have the original Altec meter, and decided it would be good to know what the scale points were, so I measured it.
I have also seen units on ebone that were forcibly used as organ donors (no meter, no transformers), so this info might be useful there as well.
I've attached my findings to this post. The short story is that it's a 200microamp movement, but the 34R shunt resistor makes it 6mA full scale.

I'd like to have a VU meter on this thing, and most natural would be to have the VU meter be the GR meter, but there's no guarantee that the scales would match, so measuring the meter was the first step.  Perhaps the simplest thing is to have two meters on the panel, one the original Altec meter, and the other a legit VU meter. Of course, now that you know the scale of the Altec meter, you could go to a meter manufacturer like Hoyt and buy a VU meter AND a matching microammeter with the correct scale supplied. If you didn't know, most meter suppliers have the means to produce custom scales for your meter. The extra cost is not as bad as you might think.

AHHH... you might be able to have the GR scale added to the VU scale, but the readings won't be what my calculated percentages are because of the meter rectifier.

enjoy!
--rick chinn

 

[MagnetoSound]

Useful! Thanks, Rick.

 

[Gearsix]

Some years ago I built a clone of the Altec 436, that now I want to modify it with EMI specs, but I am not sure about the voltage of the secondary winding of the power transformer in the EMI schematic
It is 120 Volt (as the Altec) or 240 Volt?
Thank you in advance

 

[DaveP]

Gearsix,

Some years ago I built a clone of the Altec 436, that now I want to modify it with EMI specs, but I am not sure about the voltage of the secondary winding of the power transformer in the EMI schematic
It is 120 Volt (as the Altec) or 240 Volt?
Thank you in advance

It's neither,  There is a bridge rectifier on the EMI version with a 255V DC output, so as the bridge multiplies by 1.414, its secondary must be 180Vac.  The voltages in circles are the voltages when its in compression.

best
DaveP

 

[Gearsix]

Thank you very much DaveP!

 

[rickc]

Thinking further about the VU meter vs GR meter problem with the 436/8.

I want to use a VU meter as the GR meter, but the meter rectifier likely makes it impossible because of nonlinearity at low signal levels. The next most obvious thing to do is to remove the meter rectifier, so it's no longer part of the circuit, and use the meter movement directly for GR indication. You can do this, but you'll have to develop a custom scale for GR indication (which you would have to do anyway if you left the rectifier in place).

You would need to reuse the original rectifier, as the meter scales are 'modified' to take into account for the particular rectifier's own nonlinearity, so just replacing the meter rectifier with one made of 1N34 diodes likely wouldn't work well, and using 1N914 silicon diodes would really pose problems.  Note that this is not an issue of the needle not wiggling with the input signal, it will. It's an issue of putting in a signal that makes the meter read 0VU, and then reducing that signal by 10dB, and having the meter needle realign at the -10dB mark.

Anyway, it's a nicely packed can of worms. Open at your own risk, and you should definitely have secondary containment on hand for escapees. Enough thinking about this for now, I'll just try it and find out. I could get lucky and it might just work... yeah, right... and the Easter Bunny is going to bring me a U47.

 

[Gearsix]

Just another thing!
In the EMI schematic there are some 20nF capacitors instead of the 22nF
It is the same things to use 22nF or it is better to find some 20nf capacitors?
Thank you

 

[MagnetoSound]

In the EMI schematic there are some 20nF capacitors instead of the 22nF
It is the same things to use 22nF or it is better to find some 20nf capacitors?

What is the tolerance of these cap values? Tighter than 10%?

Probably not. 

 

[Gearsix]

I am not sure about it. In my experiece the tolerance sometimes is higher than the one indicated on the capacitors
Anyway, the difference of 2nf seems to be not too big ...

 

[rickc]

in the old days, component values seemed to be at numerically convenient intervals, rather than values on the standard value tables.
While 10% resistors mostly followed the e-12 standard value sequence: 10, 12, 15, 18, 22, 27, 33, 39, 47, 56, 68, 82, capacitors were aligned on more numerically convenient numbers:
Looking at a 1967 Allied catalog, in the section for Sprague Black Beauty capacitors, I see a mix of values suggesting that maybe the company was in the process of changing from the old custom to making their parts fit the standard value table. I'll start at .01mfd and work up to 1.0mfd:  .01mf, .02mf, .047mf, .05mf, .1mf, .15mf, .22mf, .25mf, .47mf, .5mf, 1.0mf

The E-6 standard values, for 20% tolerance parts is:

10, 15, 22, 33, 47, 68

BTW, the number in the standard value designation , like E-24, represents the number of values in a decade (10x) of values, so 5% values use the E24 sequence, and there are 24 values in a decade.

In the case of the 436, I don't think it matters, and if you replace a .02mf with .022mf, you're going bigger, and any LF corner created moves downward. Now, I have found, thru tinkiering with my own 436, that there are limits to how much you can speed up the attack, and how much you can lower the threshold, and how fast you can make the release time. The penalty paid is instability, and if you do it right, the unit will sit there an oscillate. The loop is from the output stage back thru the control voltage processor (sidechain).

I see in my 436B schematic that altec used .022mf for the coupling capacitors, against a 1Meg resistor, makes for a 7.24hz LF corner. If the cap went to .02mf, then the LF corner made goes to 7.96hz. further, I see that they're .022mf in the 436B, 436C, and 438A.

 

[Gearsix]

Thank you rickc.
When I built my own clone I had some problems of oscillations that I have solved removing the input pot. I put a 2k2 pot before the first stage of the input transformer (similar to Universal Audio 175)
I'll try to put again the pot in the right position when I will modify the circuit with EMI specs.
I used a 10k:10k Edcor transformer, but I am not sure if this transformer is the right one (I read that the Altec transformer is a step up 15k:60K

 

[rickc]

to be clear, the instability was low frequency, motorboating, and I think it was related to the inherent noise level of the circuit, which was being detected by the sidechain, and then sent to the varimu stage grids, which was reducing the gain (and creating a disturbance) which was again detected by the sidechain.

Changing (raising) the threshold helped, so did lengthening the release time, and/or the attack time. It was a hot mess.

I don't see how the input pot would affect this, and I wouldn't expect it to matter whether it was an adjustable pad ahead of the transformer or a dual pot at the secondary en-route to the tube grids.

One thing that this unit needs now, more than ever, is a way to balance the varimu stage for best CMR.

 

[DaveP]

We don't realise how much of this low frequency feedback goes through the HT line and caps.

Sometimes two opposite phase stages are paired to an HT cap to cancel each other out.

At motorboating frequencies,  1 or 2 Hz, normal value electrolytics have quite high resistances.  Like rickc says, the side chain picks these up if it is sensitive to very low stuff.

Like Morgan Jones says, a regulated power supply cures it at source, but a cheap way is to stick a few 220uF and 470uF 450V caps in instead, I found this cured a low freq wobble on a WCF once.
They did not exist back in the day but are quite affordable now.

best

DaveP

 

[rickc]

Dave,

yes, it is quite possible that this was also propagated via the B+ line.  Like you said, the power supply is pitifully low on capacitance.

But like you said, a regulator would attack the problem at its source, and perhaps it would be sufficient to lower the values of the R parts of the power supply filtering and to add some zeners to overcome the supply bounce under high signal conditions.

Replacing the selenium rectifier in the doubler (haven't looked, but assuming that it is) would also stiffen things up, give a bit more B+, which we could then eat in a shunt regulator.

Yes, for sure, adding more capacitance will not hurt. Looking at the stage that feeds the varimu stage, the 10u cap is about 800-ohms at 20hz,  which is only 22dB loss at 20hz.  It's a bit better at 60hz, but only 3x in terms of the capacitor's reactance (impedance).  The saving grace for the circuit is that it's all pushpull, so it has better PS rejection than a single ended circuit. If you look at the circuit for the 438, which has a single-ended mic preamp stage, you'll see that the B+ is much more heavily filtered...

 

[Gearsix]

I don't know why, but with the dual 50k pot I have a lot of oscillations that disappear when I remove the pot, so I put it before the transformer as adjustable pad (if someone is interested I am posting a design that show it)