Bridged T Series Resistor Burning

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

gg85

Well-known member
Joined
Nov 26, 2008
Messages
121
Hey guys, I have built a 24 position 600 ohm bridged t for the output of a Gates STA build. I finally got it all up and running yesterday but noticed one of the 600 ohm series resistors on the attenuator had burnt out. I replace it and measured the voltage across it and when I'm attenuating quite a bit, or driving the input pretty high, the resistor is heating up a lot and I'm sure if I push it further again it will just burn out. There's about 65-70V across it so it makes sense that it would be getting so hot as I'm only using a 1/2W resistor, but even a 2W would burn out with that voltage. I copied my design from Friesdan's, just changing the attenuation values to suit my needs.

I can have the attenuator pulling the signal down about 46dB and it will do that fine, but if I drive the signal in a little harder with the input it will start to burn that resistor. I'm using the Hammond 1650G on the output. I don't know if I'm missing something simple here, or if I've wired it incorrectly so that it's drawing way too much through that resistor, but I also can't imagine using 2 huge series resistors in place of the 1/2W or 2W.

Here's a link to the drawing that I followed for my design: https://www.dropbox.com/s/gvilt1th1ofvu83/bridged%20t%20600%20ohm.jpg?dl=0

The issue it's the series resistor connected to the input of the attenuator the burns up. I'm using Drip's pcb for the build, and I have the "to attenuator" as my IN, "from attenuator" as my OUT, and COM connected to the pole of the 2nd deck. I've also connected COM to ground as my signal was noisy without doing this. It was a bit unclear to me as to whether it needed that but it was the only way I could get a clean output signal on my scope.

Any help is hugely appreciated!

George

 
How timely:

https://groupdiy.com/index.php?topic=70980.0

Put another fixed pad in front of it?  Are you really using the full range? 

If you're getting a hot output level with -46dB on the output and significant compression, you are indeed putting multiple watts into it. 

Have you read the full original manual and thought about the original gain stage concepts with the various fixed input and output pads?  Check that if you haven't. 
 
gg85 said:
I've also connected COM to ground as my signal was noisy without doing this.
That's why you burn this resistor! By doing this you've altered the structure of the attenuator and this resistor is submitted to too large a voltage.

It was a bit unclear to me as to whether it needed that but it was the only way I could get a clean output signal on my scope.
You need to connect the ground clip of the scope probe to the common; then you should not have undue noise.
 
COM should be able to connect to ground, or float, either way there's way too much level into this attenuator.  A Stalevel isn't designed to operate in that territory. 
 
Assuming the sSta-level has a balanced output you should connect its hot output to IN and its cold output to COM. The output hot comes from OUT and the cold from COM.

If you are measuring 70V peak to peak across one of the 600 ohm resistors that is just under 25V rms which is +30dBu which is one watt. If you really need to operate at such high levels then you need to use a 5W resistor. If you dropped the level by just 10dB then the power drops to 100mW.

Cheers

Ian
 
The other thing to remember is this type of bridged T attenuator is designed yo work with 600 ohm source and load resistances. THis means if the St-level is not already configured for 600 ohms output it will need a build out resistor to make it 600 ohms. Also you will need to add a 600 ohm resistor across the output of you attenuator unless the load it is feeding is actually 600 ohms.

Running like this, with halve the voltage across the problem resistor and drop the maximum power to 250mW.

Cheers

Ian
 
The output of a StayLevel is two 6V6, and won't "get dirty" until around 10 Watts output.

70V RMS is +39dBm which is 8 Watts which confirms that you are really hard-slamming the two 6V6es.

8W-10W into any conventional input will SMOKE it.

Or any 1W-2W resistor, of course.

So you put a pad in. If you start with 10 Watts and want to end up near 1mW, "something" has to heat-up with 9.999 Watts.

+39dBm and your maximum 24dB pad is still +15dBm, a VERY hot level for any studio.

A 16dB fixed pad will give +23dBm, and then your variable pad will deliver +23 to -1dBm, "reasonable" levels, with only 0.2 Watts of heat in the variable pad (use 1/2W parts). The 16dB fixed will be throwing 9.8 Watts, so build it with 10W or 20W resistors. Actually an O-pad only needs one big-hot resistor, the others can be modest.

For some reason Google is not finding O-pad cheaters, but Pi-pads are common and an O is just two Pi.

 
The big question is: is the 70V measurement peak to peak or rms

Edit: looking at the schematic and 6V6 data sheet I can quite easily believe it could be 70V rms. The plate voltage is 300 so there is a 600V peak to peak voltage swing available which is  212V rms. Recommended 6V6 plate load is 5K so output transformer for 600 ohms will be about 3:1 so the output could be 70V rms.

However, the Sta-level schematic shows a built in pad at the output that is variable over a 10dB to 14dB range. It has a true 600 ohm output so any 600 ohm load is going to receive 6dB less power hence the max level into the OPs attenuator ought to be  39dBu less at least 16dB or +23dBu which is a little over 10V rms. I wonder if the OPs Sta-level has the built in outout attenuator of the original.

Sta-level schematic is here:

http://www225.pair.com/audio/waltzingbear/Schematics/Gates/Sta_Level.htm

Cheers

Ian

P.S. I have a special affection for the 6V6. The very first tube amp I built when I was about 11 years old used a 6J7 and a 6V6.
 
Thank you all so much for your help, I really appreciate it! I also am aware this project has been a huge learning curve for me and I definitely have found some aspects of it way over my head. I hope it’s ok to be asking trivial questions and it doesn’t matter that I don’t yet fully understand what’s happening, although I’m aiming to learn exactly what’s going on :)

I’ve been measuring from ground to the input of that resistor with my DMM measuring AC...would that mean 0-peak?

I have no pad on the input, and am using a 100k dual log pot at this stage. I’m sending a 1k tone in at +4. I’m using a lundhall 1922 on the input, originally wired 1:8 which I’ve changed to 1:4. The thing is, I can hear the tone through the unit as I wind up the input pot, whether it’s through the tubes or transformer...I can definitely hear it, and it’s way too hot. But if I pad the input, the unit won’t compress as much, so I don’t think the issue is too hot into the 6386.

I also have a 20db pad on the output before the 600ohm attenuator I made. If I completely bypass my attenuator, the output signal is definitely way more than I need, and would ever use in a studio. I just don’t know if my issue is before the attenuator or not...as in, whether there is way too much gain being applied by the Sta, or if the issue is in the sidechain where not enough compression is being applied for the amount of signal in, therefor resulting in way too much signal going out, and how I would check that.

If the input pot was fully CW (no attenuation) and my output was fully CW (no attenuation) and I had no input pad with the input tranny wires 1:4 but had the 20db pad on the output, in theory the unit should be compressing around the 30db mark and my output should be within a useable range, am I correct in my thinking?

And again thank you all for your help!
 
Ask away. As they say, the only dumb question is the one you don't ask Plus this ain't gearslutz.

Here's a link to the Sta-level manual:

http://www.realoldiesradio.com/docs/StaLevel/Sta-Level_Manual.pdf

It says with all the standard pads in circuit the gain is 35dB. With them all out it is 62dB - that's enough to make it a mic pre.

It also says the ratio is about 3.3:1 and the threshold is about +10dBm output.

So start with a -25dBu input and you should get about +10dBu out and no compression. If you increase the input by 30dB to +5dBu,  the output should increase by less than 10dBu to just under +20dBu and the compression meter should read about 20dB of gain reduction. If you look at this schematic:

Sta-Level.jpg


You should see the annotation for control voltage versus gain reduction so in the case above the voltage should be around 19 volts.

Cheers

Ian
 
A bit of an update...I had been hooking the attenuator up incorrectly and had the 20db pad bypassed unintentionally so my attenuator was getting the full level through it. I've now got a 10db pad with 5W resistors just to be safe, and the attenuator is working great with no resistors heating up and getting to maybe a 1/4W through the series resistor. I've also got the input tranny wired 1:4 so there's a little more play with the input attenuator. The other factor that might have contributed to such hot output levels is the 6600:500 output. I was considering using the 250 ohm tap for a bigger step down and rebuilding the attenuator to suit, but I don't think i'll need to do that now! Thank you so much for all your help. I had seen those CV's on the schematic but couldn't read it properly and thought the V was a % (which makes no sense) and was lost for checking actual compression. So now the unit is working as it should, which is great :) The one and only obstacle to overcome now is that I am using Simpson 142 VU's for the metering, which are not DC. I love the look of the meters, and I'm going to make my own scale so they're useable, but I have one last question regarding this build. Is there a mod on either the Sta or the meters that I could experiment with to make the needle swing a little more at lower compression settings?
 
Awesome.

Why not find a Simpson 1mA meter in that form factor?  It'll be of less $ value than the VU, and work properly right out of the gate.  They're not growing VU's like that on trees anymore. 

A proper 1mA meter may swing more responsively than a VU which has a movement damping spec.  Outside of that thought, peaks are always invisible on compressors like this, the slight movement is more an indication of threshold and action.  You'd need a peak reading circuit added to get the other extreme. 
 
I’m not sure if I’ve got this completely wrong, but the meters I have are essentially DC meters with a rectifier and resistors for VU use right? They seem to be 200ua FSD. Is it possible at all to just remove the rectifier and resistors and use them as dc meters? Is there any way to make them work like a 1mA would? I’m not too fussed that I won’t be seeing peaks as it’s always more of an indication of compression times and a general idea of what’s happening. Plus these meters just looks so good!
 
gg85 said:
I’m not sure if I’ve got this completely wrong, but the meters I have are essentially DC meters with a rectifier and resistors for VU use right? They seem to be 200ua FSD. Is it possible at all to just remove the rectifier and resistors and use them as dc meters? Is there any way to make them work like a 1mA would? I’m not too fussed that I won’t be seeing peaks as it’s always more of an indication of compression times and a general idea of what’s happening. Plus these meters just looks so good!

All moving coil meters are dc meters. All VU meters are dc meters with a rectifier attached so they can read ac. What makes a VU meter a true VU meter is the meter ballistics and the rectifier characteristics.

Your 200uA dc meter can be made into a 1mA meter by adding a shunt resistorin parallel. The shunt needs to be designed to take 800uA of a 1mA input current so only 200uA flows in the meter. To work this out you really need to know the dc resistance of the meter. As it takes four times the current, the shunt resistor needs to be one fourth the dc resistance of the meter.

Cheers

Ian
 
You'll get faster ballistics out of a proper 1mA meter. 

You'll be less angry at yourself the next time you have to spend big money on some real VU's and remember what happened to these.  FWIW.....  If you use them, as you said, you don't really need a scale, there are several compressors with no scale marked. 
 
Thank you so much guys! I’m definitely going to try these meters first and see how they go but if it’s not great I’ll find some 1mA meters to do the job. What would be the best way to measure the resistance of the meters? Is it as simple as measuring across the meter (no rectifier or resistors) with a DMM?
 
The best way would be to use a resistor and a variable power supply. Start with 100K resistor. Wind the voltage up slowly, dont let the meter pin. Once you have full scale, measure the voltage across the meter and use ohms law to calculate resistance.
 
Back
Top