In my case with the Altec model 9's they are a three way speaker with low, mid, and high frequency. An 800 hz horn might be nice or maybe a 500 hz horn a little better to combine with the proper low frequency speaker and proper passive crossover. Maybe a really nice setup would be to use the 6420 on highs only and biamp with a beefier amplifier to drive the lows. Naturally one would need to incorporate a nice electronic crossover.

I don't know what the amplifier or I should say amplifier pieces will produce for power when I re-do the amplifier with a decent driver circuit and run the amp Class AB in triode mode.

Oh, someone has two of the 6420's on ebay at the moment.

Rather confusing having multiple people posting about having the same amp. The last post was directed at Daniel P

Burntfingers, that sounds like a great idea. You could use a DBX electronic crossover (i use it in my system and it is great) then use a Yamaha P2500S amp for the lows.

This Eminence driver http://www.parts-express.com/pe/showdet ... er=290-442 will be flat to 16 KHz when used with the ALTEC 811B horn. Crossover at 1.2-2 KHz. Also you would need to insert a 1156 bulb in series with the driver as it makes the horn sound not quite so harsh.

quote:
This Eminence driver http://www.parts-express.com/pe/showdet ... er=290-442 will be flat to 16 KHz when used with the ALTEC 811B horn. Crossover at 1.2-2 KHz. Also you would need to insert a 1156 bulb in series with the driver as it makes the horn sound not quite so harsh.


I can get my hands on Altec drivers thru work. Actually the bulb will serve to limit the power handling also. Have used a 60 watt car bulb in series with HF drivers. Light turns on hard at 60 watts and serves as a limiter. No more broken burned out voice coils on those expensive diaphrams.

Anyone have any specifications on the power transformer or output transformer off the 6420?

If I can figure out how to post a picture I will put one up of the Model 400 Tone Cabinet which has the 6420 amplifier and the tone cabinet schematic.

I've been using my 6420's for quite some time and they've been doing well. The only issue I've been having is a slight amount of red plating on the output tubes. One of the two will ever so slightly red plate. I decided to take a look around and first decided to measure the drop across the cathode bias resistor which is a 240 ohm 10 watt wirewound. I measured 30 volts. I then measured the plate voltage by measuring the between the cathode and the plate and I measured 383 volts. I measured the screen voltage as 373 volts.

According to the bias calculator that I've been using (http://www.webervst.com/tubes1/calcbias.htm) it seems like the tubes are biased much too hot. Considering the 30 volt drop across the cathode resistor and the 383 volts on the plate I'm well over 70% plate dissipation.

The other worrying issue is that the screen voltage is much too high. According to this schematic (http://img843.imageshack.us/img843/5240/fullschem.jpg) the screen voltage should be roughly 300 volts whereas both of my units have in the range of 375 volts on the screen. The screen voltage is obtained from the B+ line and is then reduced by R9, a 1700ohm 50 watt wirewound resistor, before being sent to the screens. The B+ before the resistor is 383 volts and after the resistor is 373 volts. I would expect a larger drop than this, correct? How are they managing to get it down to ~300 volts?

I've measured both the cathode bias resistor and the screen voltage dropping resistor in both amplifiers and they're both very close to spec.

Are you running a field coil between pins 1 and 7 of the #2 speaker socket? If not a 6K resistor should be used.

I am not. What wattage would be needed and what function does it perform?

check this out

http://dynaco2forum.forumsmotion.com/

I have a 6420 section that I started

The field coil drops the voltage and filters the B+ supply. Inserting a 6K resistor of 30watts to try will give you an idea of how big/wattage the 6K resistor will need to be. Old speakers didn't have a permanent magnet so the field coil was the magnet structure.

If you do use a resistor the filter capacitor that connects after the resistor will need to be increased in value.

Okay, that makes sense. How much would the filter capacitance need to be increased? I've also heard that using a choke would be a better idea due to the in built filtering properties. What sort of specification choke would be needed?

Why not build a bias circuit so you can adjust each output tube?

Then convert the amp to triode operation.

Convert it to fixed rather than cathode bias? Perhaps at a later date. If installing a good quality choke will allow for the amp to operate at its proper voltage I'd prefer to take that route first.

Yes you could use a choke then use a resistor in series with the choke such that the dc resistance of the choke and resistor equal or come pretty close to the field coil resistance.

Okay, so 6k total resistance? Would 7-10Hy work well or would I need more inductance? This is a....capacitor input power supply, correct? Therefore I wouldn't need much more than 50-75ma of current handling capacity?

I suppose that would be enough inductance. Could probably put the inductor after the resistor which would increase its effectiveness.

Also, if I decided to instead increase the filtering of the filter capacitor instead of using a choke, how much higher would I need to go over the stock value of 40uf?

Not exactly sure, but maybe 60 uF would work and given the capacitor will be after the resistor you could even use 200 uF with no ill effects on the power supply.

I just finished working on my 6420's put in a bias circuit adjustable per output tubes and made the amp triode.

Now I am wanting to make a screen supply if I desire to use it in pentode mode. How do I use 4) 0C2's in series to achieve 300 vdc for the screen off the main supply. Suggestions would be appreciated.

You'd be better off with a pair of OD3's instead of four OC2's. The fewer tubes in a VR stack, the better.

The standard practice is to feed the VR stack from the unregulated supply rail by way of a limiting resistor. The load (in this case the 6420 screen grids) is connected in parallel with the VR stack, and the two work in a "seesaw" arrangement, i.e., as load current increases, current through the VR stack decreases.

For stable operation, a minimum VR tube current of 5 to 10 ma is required. The maximum permissible current with most types is 40 ma. Therefore, the load current draw cannot exceed 30 to 35 ma if the voltage is to be stabilized over a range from zero to maximum load current.

The value of the limiting resister must lie between that which just permits minimum VR tube current to flow, and that which just passes the maximum permissible tube current when there is no load current. The latter value is generally used.

It is given by the equation: R = 1000 (Ex - Er) / I, where R is the limiting resistance in ohms, Ex is the source voltage, Er is the rated voltage drop of the VR stack, and I is the maximum VR tube current in milliamperes (usually 40).

With two or more VR tubes in a stack, current-limiting resistors of 50 to 1000 ohms apiece should be inserted between the cathode of each tube and the plate of the next tube below it, in order to assure that no tube carries more or less than its fair share of the load. With two VR tubes in a stack, you will need only one such resistor; with four tubes, you will need three, and this is one reason why you should use as few VR tubes as possible.

The wattage rating for all the resistors can be worked out by simple Ohm's Law calculations, based on a maximum current draw of 40 ma.

Larry

Went with 3) 100volt zeniers instead of the VR tubes. Sounds really strong now better than it did in triode mode.
Added a bias transformer and circuit so that I could adjust the bias on each output tube. My bias voltage going to the grid resistors (92K) changed from -42-45 to -24 -25 range is this correct? running plate of around 470 and 45mA per output tube.

Of course it sounds stronger than it did; you've at least doubled your output power.

As for the bias, it's not such an easy matter to call. You have the choice of biasing the tubes in class A (which wastes a lot of power in heating the hell out of the tube for little or no tangible gain) or biasing them in the class AB1 range, which is more efficient in operation, less wasteful of power, and much less stressful on the tubes.

I personally prefer deep class AB1. I like to baby an output stage so it'll last. It won't cost me anything I'm apt to miss, and I certainly won't miss the expense of retubing after a pitifully small number of hours. But it's up to you; just try not to stray far from the book, if you want things to last.

Good luck with your project.

Larry