Running a tube amp without a load causes the voltage at the plate of the output tubes to run away. I have seen voltages over 800 volts on a 5 watt SE amp. With the voltages available in this amp a fireball could erupt inside your OPT instantly if the voltage rises above the breakdown voltage of the insulation in the OPT.
A tube amp will work into a short circuit far bettet than an open. If you are testing your amp, and a load is not available, short the output leads. This will prevent an arc from happening. Running any power through the amp into a short will make your output tube unhappy, but usually won't blow anything.
Radio Shack used to sell an 8 ohm 30 watt "non - inductive" load resistor. I have put 80 watts into these without failure. For good measure I suppose you could put 4 in series parallel to make an 8 ohm 120 watt load. The store near me does not have these any more.
A tube amp will work into a short circuit far bettet than an open. If you are testing your amp, and a load is not available, short the output leads. This will prevent an arc from happening. Running any power through the amp into a short will make your output tube unhappy, but usually won't blow anything.
Radio Shack used to sell an 8 ohm 30 watt "non - inductive" load resistor. I have put 80 watts into these without failure. For good measure I suppose you could put 4 in series parallel to make an 8 ohm 120 watt load. The store near me does not have these any more.
I got a couple of those dummy load resistors from the Rat Shack right up the street when doing R & D on the 807 amp. Found it most convenient to clamp it to some aluminum angle stock as a heat sink as the resistor gets plenty warn with 25W going into it.
My rat shack carries monster cable, printer ink, and superflous warranties these days.
Fortunetly Norwood is blessed with a real electronics supply Debco, but unfortunetly it is threatened with eminent domain. It looks like they arent restocking the way they used to, but I like checking thier specials. The US now feels that buldozing properties that pay lower taxes with ones that will pay higher taxes is a legitimate use of eminent domain. I have watched my city try to use eminent domain to put in a Walgreens pharmacy across from a CVS pharmacy, if they did it now they would likely have succeeded. Property rights are a thing of the past for those who arent big business or connected.
In this particular case I wouldnt worry about whether the dummy load is non-inductive, its not called a voice coil for nothing.
Fortunetly Norwood is blessed with a real electronics supply Debco, but unfortunetly it is threatened with eminent domain. It looks like they arent restocking the way they used to, but I like checking thier specials. The US now feels that buldozing properties that pay lower taxes with ones that will pay higher taxes is a legitimate use of eminent domain. I have watched my city try to use eminent domain to put in a Walgreens pharmacy across from a CVS pharmacy, if they did it now they would likely have succeeded. Property rights are a thing of the past for those who arent big business or connected.
In this particular case I wouldnt worry about whether the dummy load is non-inductive, its not called a voice coil for nothing.
I was experimenting with a screen drive P-P 6AV5 amp. SY made me do it! I expected about 30 watts, but I kept turning up the drive voltage and the power supply, and more power kept happening. After I figured out that I could get 80 watts out of this thing, I should have dug up the 500 watt resistors that I got off of Ebay a few years ago, but I was lazy. I just got a coffee mug full of cold water and dunked the load resistor in. No don't try this at home. I cranked 80 watts for several minutes with the room lights off to check the tubes carefully for glowing plates and screen grids. They were happy. The water was beginning to boil. I changed the water and ran frequency response tests. The load resistor lost its lettering, but is otherwise unaffected.
The 500 watt resistors are inductive enough to affect frequency response tests by about 1 or 2 db at 20 KHz.
Bias supply
you have no choice but low impedance IF you want A2.
I have been looking at a number of Japanese designs. They use regulation from adjustable three terminal regulators.
My friend says brute force is the go and hang the heat and waste. This means high bleeder current and low value Rs in the divider.
Just make sure your last cap is large. .i.e. e.g. >200uF. This will at least give you A2 headroom for peaks.
Regs might be the most sane option but look dodgy to me as low voltage devices stacked up floating on high voltage.
If they blow due to fault ..... you are in real trouble!!!
Maybe safer to stack some high voltage zeners and big cap.
I suppose a high impedance divider in parallel could hold the bias in case of failure.
Cheers,
you have no choice but low impedance IF you want A2.
I have been looking at a number of Japanese designs. They use regulation from adjustable three terminal regulators.
My friend says brute force is the go and hang the heat and waste. This means high bleeder current and low value Rs in the divider.
Just make sure your last cap is large. .i.e. e.g. >200uF. This will at least give you A2 headroom for peaks.
Regs might be the most sane option but look dodgy to me as low voltage devices stacked up floating on high voltage.
If they blow due to fault ..... you are in real trouble!!!
Maybe safer to stack some high voltage zeners and big cap.
I suppose a high impedance divider in parallel could hold the bias in case of failure.
Cheers,
Just out of curiousity....
WHat is the input sensitivity of this amp????
Just doing quick and dirty analysis...you first tube, 7193 in it's first gain stage will only make a Gain of 15 at best conditions...
The 300B into a 1:1 transformer, unloaded will only get you a gain of 4 at best case.... WHat is the value of the damping resistor on the secondary side of the driver transformer???
If this value is too small it will further reduce the Gain...
SO far I am getting a Gain of 60 by the time you drive the 845??
And that is if everthing is better than perfect...
So input range of 2V RMS will result in 120V signal at 845??? That is ideal case...
Am I missing something??
Chris
WHat is the input sensitivity of this amp????
Just doing quick and dirty analysis...you first tube, 7193 in it's first gain stage will only make a Gain of 15 at best conditions...
The 300B into a 1:1 transformer, unloaded will only get you a gain of 4 at best case.... WHat is the value of the damping resistor on the secondary side of the driver transformer???
If this value is too small it will further reduce the Gain...
SO far I am getting a Gain of 60 by the time you drive the 845??
And that is if everthing is better than perfect...
So input range of 2V RMS will result in 120V signal at 845??? That is ideal case...
Am I missing something??
Chris
Bryan said:
Typically you design a power amp to have a input sensitivity of something like .5V RMS to 1V RMS for full power output..
I tend to use 1V RMS input for full power output...
The line stage is usually not expected to be running at full gain with pot wide open... I prefer the pot somewhere like 5 to 7 for full output from the power amp...or 1V input..
You want to keep your S/N ration to some reasonable limits...
A 6SL7 with both sides paralleled may do the trick for your 1st stage in your amp....
Chris
If you use a 6SL7 in the 1st gain stage....parallel both triode sides of the tube...
Change the plate resistor to 100K then change the 300B grid bias resistor from 100K up to no less than 220K, preferably 470K if allowed...
Leave everything else the same...your cathode resistor should get 1.5V across it...and the plates should be at 150V...you would need to adjust the supply dropping resistor for this stage to obtain 300V still...
Now you have a 1st stage Gain of 53 ...
Your 300B stage gain is roughly 3.9......
So now you will produce 200V at the grid of the 845 with roughly a 1V input.... Now you can keep your line-level volume control at a reasonable spot to keep the S/N ratio down..
There are nicer choices for input valve..but i figured you wanted something along these lines...
Chris
Change the plate resistor to 100K then change the 300B grid bias resistor from 100K up to no less than 220K, preferably 470K if allowed...
Leave everything else the same...your cathode resistor should get 1.5V across it...and the plates should be at 150V...you would need to adjust the supply dropping resistor for this stage to obtain 300V still...
Now you have a 1st stage Gain of 53 ...
Your 300B stage gain is roughly 3.9......
So now you will produce 200V at the grid of the 845 with roughly a 1V input.... Now you can keep your line-level volume control at a reasonable spot to keep the S/N ratio down..
There are nicer choices for input valve..but i figured you wanted something along these lines...
Chris
cerrem,
Thanks for the input. Much appreciated!!!
The amp is going to be a work in progress. The tube line-up will presumably need a little "tweaking" as I progress. The most important aspect of the design is that I learn a bunch, and have some fun tweaking. I want to get it opperational, and then begin experimenting.
That said, what would you think to be more "exciting" for the front end?
I was thinking about the possibility of using a CCS loaded 76, but this has the same low gain...
What would be on your list of favorites?
Thanks for the input. Much appreciated!!!
The amp is going to be a work in progress. The tube line-up will presumably need a little "tweaking" as I progress. The most important aspect of the design is that I learn a bunch, and have some fun tweaking. I want to get it opperational, and then begin experimenting.
That said, what would you think to be more "exciting" for the front end?
I was thinking about the possibility of using a CCS loaded 76, but this has the same low gain...
What would be on your list of favorites?
Hey Guys,
So here is a strange one....
I've poswered up the amp, and everything seems to be working fine.
Today, I applied some signal and no music. So, the first thing to do was to see if I was amplifying the signal at the various stages.
The input stage seems to work, and is swinging voltage on the drig of the 300B. I also have voltage swing on the grid of the 845, but no music cometh from the speakers. Admittedly, I have not fooled around measuring near the 845, but I know I have plate voltage (1000VDC) and I know the tube is passing current (70mA) and I know the negative bias is working (-145VDC). Thoughts.
And another even stranger thing.... This I noticed both with and without the 845 HV applied. When I turn up to volume of the source (my preamp) I can hear music coming from the IT transformer. How is this happening???
Thoughts?
Thanks,
Bryan
So here is a strange one....
I've poswered up the amp, and everything seems to be working fine.
Today, I applied some signal and no music. So, the first thing to do was to see if I was amplifying the signal at the various stages.
The input stage seems to work, and is swinging voltage on the drig of the 300B. I also have voltage swing on the grid of the 845, but no music cometh from the speakers. Admittedly, I have not fooled around measuring near the 845, but I know I have plate voltage (1000VDC) and I know the tube is passing current (70mA) and I know the negative bias is working (-145VDC). Thoughts.
And another even stranger thing.... This I noticed both with and without the 845 HV applied. When I turn up to volume of the source (my preamp) I can hear music coming from the IT transformer. How is this happening???
Thoughts?
Thanks,
Bryan
A singing transformer is usually a sign that there is no loading on it's secondary as the audio modulation voltage across it's primary rises. The effect making the sound is called magnetostriction. With the amp unplugged and capacitors discharged I'd check the continuity to the grid pins on the 845 sockets back to the IT secondary and the wiring here. You may have overlooked something.
No loading on the secondary of the IT. That is probably it. Looking back at my schematic, there is a resistor connecting both legs of the IT secondary. Currently that is not present. I was under the impression that this resistor is not absolutely necessary, and should be used to tune out any ringing on the secondary.... Does the 845 itself not load the IT secondary???
I am not sure what value resistor to use in this applicaiton. What are the rules of thumb for determining the value, and if it is absent, would this explain my singing IT and no music coming from the amp OPT???
Thanks,
BK
I am not sure what value resistor to use in this applicaiton. What are the rules of thumb for determining the value, and if it is absent, would this explain my singing IT and no music coming from the amp OPT???
Thanks,
BK
Have you posted your schematic? I don't know what you've wired up here. In A2 you need the IT secondary connected with DC continuity from master B+ supply negative ( usually chassis ground), through the positive grid bias supply, through the IT secondary, to the 845 grid. Is this a SET or P-P amp? The R you speak of placed across the IT secondary is probably placed there to equalize the frequency response of the transformer. Extend the bass and so forth. Leaving it out will not cause the amp to not function.
I just re-read your question. You mention both legs of the IT secondary. That implies P-P to me. If there are two separate secondary windings on the IT, one for each output tube, and there was supposed to be a R between the two inner connections to the windings then yes, it has to be present. It may be employed to create the grid bias but I''d need to see the schematic. Is it possible there is to be a R from each leg at the IT center to ground? The two inner ends of the IT secondary windings absolutely have to be connected somewhere or you won't have audio action. I dunno how you are getting what appears to be proper current through the 845's now if the grids are floating.
I just re-read your question. You mention both legs of the IT secondary. That implies P-P to me. If there are two separate secondary windings on the IT, one for each output tube, and there was supposed to be a R between the two inner connections to the windings then yes, it has to be present. It may be employed to create the grid bias but I''d need to see the schematic. Is it possible there is to be a R from each leg at the IT center to ground? The two inner ends of the IT secondary windings absolutely have to be connected somewhere or you won't have audio action. I dunno how you are getting what appears to be proper current through the 845's now if the grids are floating.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.