Heathkit Cheyenne questions

[NZ0T]

New member and poster here.  I am mostly a CW op but I recently started to re-create my Novice station circa 1982 of the Heath Cheyenne/Comanche twins.  I picked up a Cheyenne a couple weeks ago that I have been operating a bit on CW.  Not tried AM yet due to lack of a high impedance mic. Please forgive my dumb questions but I'm just not up on vintage tube rigs these days.  I am using the HP-23B PS with the rig.   With the function switch in grid position peaking the grid with the drive control sends the meter needle against the hard right stop.  I tried adjusting the drive down to 3 ma with the drive and tuning normally otherwise but I end up with about 90 watts out.  If I tune the grid to full - as much as I can with it going full right - I get the normal 65 watts or so out and the rig appears to operate normally with a good note on CW.  My worry is that even with that seemingly normal output that the 6146 will smoke if that grid indication is correct as that tube is prone to high grid current.   Could this be a metering problem?  I know the HP-23 has a higher high voltage than the HP-20 but I have seen several people post that the MT-1 will work fine with the HP-23's higher voltage.  I do have it set to 300 volts on the low end.  So, any ideas on what to do would be appreciated.

The other issue is that I am measuring about 250 volts across the key terminals instead of the 90 volts the manual says should be there.  I blew my trusty 30 year old MFJ-401 keyer trying to use it with the MT-1 - should have checked the voltage first!

Any help greatly appreciated!
73 Bill NZ0T

[WQ9E]

Bill,

The grid of the 6146 definitely doesn't like high grid current and anything of 3.5 mils is too much, 2.5 is sufficient.  Running with the grid meter pinned is going to result in rapid tube degradation if it hasn't already, you will find the grid develops secondary emission and plate current will creep as the tube gets hotter.

You should be able to reduce the loading to get the rig down to rated power.  The HP-23 will work but the B+ is higher and the MT-1 won't load it down as much as the HW-100/SB-100 series it was meant to work with so it should be loaded to a lower plate current than what is stated in the manual.

[AB2EZ]

Bill

1. You don't need a high impedance microphone to work with the Cheyenne. Any microphone that doesn't require phantom power will work fine. E.g. any dynamic microphone.

2. The grid drive control should be adjusted to produce the recommended grid current (probably 2.5mA). It should not be adjusted for maximum grid current.

3. If you want to check the accuracy of the grid current metering, measure the resistance of the grid metering shunt resistor (with the power off, and the meter switch in the plate position... so that the meter is not in parallel with the grid metering shunt resistor). As per the schematic, the value should be around 32 ohms. The is the last resistor in the chain of components between the 6146's grid pin and ground. All average grid current flows through this path. The value is not critical. Any value within 5% of 32 ohms is okay for the purpose at hand.

4. If you have a 50 ohm output load, and if the loading control is adjusted so that the plate current reading (the meter is actually measuring the cathode current; which includes the screen current) is less than 125mA when you tune for a dip in the plate current, then the RF output power will be around 40W. Remember: the electrical input power to the tube will be around 600V x 125mA = 75W. The RF output power has to be less than that... and is typically 60-75% of the electrical input power to the final amplifier tube when operating CW.

5. As to the across the key jack: I am not surprised that it is a lot more than 90V.

Good luck
Stu

[NZ0T]

Thanks for the tips guys.  I'll check the resistor values tomorrow.  For now I'm making sure grid is at or below the recommended in the manual 3 ma and I'll load it down to about 50 - 60 watts out.  The MT-1 manual does say to peak the drive and then after tuning the grid should be about 3 ma with the plate at 150 but I'm seeing the plate at around 100 ma at 60 watts out and just under 3 ma drive.

Maybe I'll try a mic tomorrow too.

[AB2EZ]

Bill

There must be an error in your RF power measurement.

With 600V of plate voltage and 100ma of cathode current, the electrical input power to the 6146 is less than 60W. You cannot have 60W of electrical input and 60W of RF output.

Stu

[WQ9E]

Stu,

I would agree his RF output meter is reading high but he also has far more than 600 volts on the final since he is using a HP-23 series supply instead of the UT-1 and his supply is going to produce better than 750 volts @ 100 mils of load if its filter caps are in good shape.

[AB2EZ]

Agreed:

The electrical input power is the actual 6146 plate voltage x the actual 6146 plate current. The actual 6146 plate current is roughly 12.5-15 mA less than the cathode current, since the cathode current is the plate current + the screen current + the grid current. In CW mode, the screen current will be around 10-12 mA, and the grid current should be 2.5-3 mA.

Stu

Stu,

I would agree his RF output meter is readings high but he also has far more than 600 volts on the final since he is using a HP-23 series supply instead of the UT-1 and his supply is going to produce better than 750 volts @ 100 mils of load if its filter caps are in good shape.

[NZ0T]

Yes I'm using a good HP-23B that produces +820 volts at no load.   My meter is an Elecraft W2 and is accurate.  My apologies if I failed to mention the higher voltage.

[NZ0T]

I had a chance to check the resistors this morning.   the 27K and 31.62K in the DC voltage divider seem to be OK.  The 27K reads 28.5 and the 31.62 reads 31.4 so they are pretty close.  However the 27K from pin 6 of the final to ground was high at about 41K.  Luckily I had a 27K 1 watt in my junkbox and I put it in.  The voltage across the key went down slightly to about 230 volts.  The grid indication has not changed.  I should have mentioned that the grid reading and tuneup do appear to work properly on ten meters and pretty close on 80 but 15, 20 and 40 exhibit the high grid indication.

The final grid choke appears to be fine as it is not shorted.

Now what? 

[AB2EZ]

Bill

1. With respect to the grid current:

In this design, there is only one control to adjust the grid current (unlike in a Johnson Ranger, where there is a control to resonate the tuned circuit at the output of the driver tube, and there is a separate control to adjust the screen voltage on the driver tube).

If you adjust that single control for maximum drive current, then you will see a much lower maximum drive current on the higher frequency bands. The driver is also acting as a frequency multiplier on the higher frequency bands... and, on the higher frequency bands, you are resonating the tuned circuit at the output of the driver tube to a harmonic in the (non-sinusoidal) plate current waveform of the driver tube. This means that the amplitude of the voltage waveform across the driver's output circuit (at resonance) is lower on the higher frequency bands... and this results in less 6146 grid current on the higher frequency bands.

Not withstanding what the manual says, after you tune and load the 6146 output tube, you should back the "drive" control away from resonance to obtain no more than the maximum rated grid current of 4ma (preferably the recommended grid current of 2.5-3 mA)

2. With respect to the voltage across the key jack:

The voltage that you are measuring implies that there is approximately 9mA flowing through the 27k ohm cathode biasing resistor when the key is up.

This voltage "cuts off" (reduces to a very small value) the cathode currents of the 6146 and the 5763 when the key is up.

The value of this resistor has only a minor effect on the voltage... since the current flowing through it on key up will decrease if the value of the resistor is higher.

It is a little puzzling as to why the voltage required to "cut off" the cathode current in both tubes is as high as what you are measuring.

The tube requiring the higher voltage to cut it off will dominate.

As an experiment... remove the 5763 from its socket, and then measure the voltage across the key jack.

Let us know what you observe

Stu

[N8ETQ]

 Bill,

   Agreed, I can't see it either.  Being a "Greifkit" one must
also make sure the good components are installed in the proper
position. Tube socket voltage checks may be helpful. You have to
almost trace it out, 4 hours of Verification may save 12 hours
of "Abstract" troubleshooting.  Hey, maybe the assembler
liked to take a "Sipp" now and again!

GL

/Dan

Bill

It is a little puzzling as to why the voltage required to "cut off" the cathode current in both tubes is as high as what you are measuring.

The tube requiring the higher voltage to cut it off will dominate.

As an experiment... remove the 5763 from its socket, and then measure the voltage across the key jack.

Let us know what you observe

Stu

[NZ0T]

The voltage across the key is 232 with the 5763 in and 78 with it out.

[N8ETQ]

ouch...

[AB2EZ]

Bill

Okay

That confirms that the problem is in the 5763 driver circuit... but not necessarily with the tube itself.

1. Remove (again) the 5763 from its socket

2. With the key up (open), measure (carefully) the following:

A. The DC voltage between the 5763 grid pin and ground. (Measure this at the end of the 1.1mH grid biasing choke that is opposite the end where the 1.1mH grid biasing choke connects to the grid pin).
B. The DC voltage between the 5763 screen pin and ground.
C. The DC voltage between the 5763 plate pin and ground.

Let us know what you find.

I suspect that, under key up conditions (or with the tube removed) there is positive voltage on the grid pin... due to insufficient DC resistance across the 5763's 47pF grid input coupling capacitor or conducting debris somewhere on the socket ... maybe between plate pin 1 and grid pin 8. Under key up conditions, there is no grid current into the 5763; and leakage between something at 300V and the grid pin could positively bias the grid voltage. This would cause the cathode voltage (the voltage across the open key) to be much larger than normal.

Stu

[AB2EZ]

It is also possible that the large cathode-to-ground voltage required to cut the 5763 off is normal for this design.

In this transmitter, the VFO and the following buffer stage remain "on" under key up conditions. Therefore, under key up conditions, there is a signal with large positive voltage swings (maybe as high as 200V) driving 5763 grid.

In order to cut off the 5763, the cathode-to-ground voltage would have to about 50-75 volts higher than the maximum positive value of that grid input signal.

That would explain the large voltage across the key jack when the 5763 is plugged into its socket.

Stu

[NZ0T]

I'll try and get those voltages checked tomorrow night after work.  Got to figure out the pin numbers lol.   FYI the MT-1 manual says "approximately 90 volts will be present at the key terminals under "key up" conditions".

It is also possible that the large cathode-to-ground voltage required to cut the 5763 off is normal for this design.

In this transmitter, the VFO and the following buffer stage remain "on" under key up conditions. Therefore, under key up conditions, there is a signal with large positive voltage swings (maybe as high as 200V) driving 5763 grid.

In order to cut off the 5763, the cathode-to-ground voltage would have to about 50-75 volts higher than the maximum positive value of that grid input signal.

That would explain the large voltage across the key jack when the 5763 is plugged into its socket.

Stu

[AB2EZ]

Bill

The manual was probably written by a summer intern.

If the there is an operating mode in which the 6AU6 VFO and the 6CL6 buffer amplifier are turned off, and where plate and screen voltage to the 5763 and the 6146 are still applied... then the voltage across the open key jack will probably drop to the range specified in the manual. I doubt there is such a mode.

But in normal CW operating mode, this transmitter does not turn off the 6AU6 VFO or the 6CL6 buffer amplifier on key up.

To test this:

Plug the 5763 back in, and remove the 6CL6. Then measure the voltage across the open key jack.

Stu

[AB2EZ]

Bill

If you confirm that removing the 6CL6 (leaving all other tubes in place) reduces the key up voltage across key jack to the value given in the manual... the following modification should keep the voltage across the open key jack from increasing when you reinstall the 6CL6:

1. Obtain four (4) diodes of type 1N4004 or 1N4007.
They cost a few cents each... and even Radio Shack usually sells them.

2. Configure the four diodes in series... being careful that they all point in the same direction (the cathode of the first diode in the series chain is connected to the anode of the second diode in the series chain, etc).

The purpose of using four (4) diodes in series is to obtain a lower capacitance (not to obtain a higher peak inverse voltage rating). A single diode has a capacitance across it of about 15pF. A chain of four (4) diodes in series has an end-to-end capacitance of around 3.75pF.

3. Solder the cathode end of the series diode chain to ground. Solder the other (anode) end of the series diode chain to the grid pin of the 5763 socket. Check that you have done this correctly, and that there are no inadvertent short circuits between the diode string and nearby components.

Note that the cathode end of each diode is the end with the grey band.

This set of four (4) diodes in series will "clamp" the positive peak of the grid voltage waveform to ground... just as the 5763's grid-to-cathode diode does on key down. We usually call this "self-biasing" because clamping the positive peak of the grid voltage waveform to ground causes the DC (average) grid voltage to be negative with respect to ground.

On key down, the added diode string is redundant, and will have essentially no effect. On key up, the added diode string will keep the voltage across the key from surging upward.

Stu

[AB2EZ]

Bill

Any update?

Stu

[NZ0T]

Hi Stu,

No update except that I found the matching Comanche receiver and UT-1 power supply that will be here Friday.  Will need a bit of work but should be interesting to see if having the correct 600V supply makes any difference.  I have been using the Cheyenne on CW a bit keeping the indicated grid at 3 ma or below and the power at around 60 watts.  No problems so far.

[NZ0T]

Got the MR-1 and UT-1 power supply yesterday.  Once I got the UT-1 going I tried it with the MT-1.  Grid still is high but reducing it to 2.5 - 3 ma and tuning resulted in the proper plate of about 150 mils and around 60 watts out which is what it should be.

The MR-1 works pretty well though the AGC seems fast on SSB and CW.  I did manage to find a foreign broadcast station to listen to on 40 meters AM this morning and it sounded very good.  Mostly it needs alignment and, Deoxit and maybe a new tube or two.  The big job will be re-stringing the dial string.  The rotating drum sting is fine but the one for the frequency indicator is broken.  Looks like I will need to take the faceplate off to do that.

73 Bill NZ0T

[NZ0T]

New update.  Got the twins up and running well on CW.  Got the dial string on the MR-1 re-strung which took me three hours!  Not much room to work but I got it done.  Spent some time aligning the MR-1 as much as I could without much test equipment and it was pretty close.   I did change the wiring so that the 300V B+ is always running to the MR-1 so I could hear the TX in the RX for sidetone on CW.  Guess I'll have to rig up a switch to defeat that when I get around to trying AM - that IS what this forum is about!

The AVC is too fast and I'm not sure if it's a bad 6T8 which is the tube in that circuit or a cap that has gone bad/changed value.  Will look at that soon.  I have found that if I don't ride the RF gain and run it too wide open I get noise and distortion which may just be how this old rig is. 

But I'm having fun with it!

73 Bill NZ0T

[NZ0T]

Latest update.  It was suggested to me in another forum that adding a larger cap in the AVC circuit would slow the AVC down but when I try that CW signals chirp unless I turn the RF gain way down so maybe I'll just live with the fast AVC/AGC.

I plan to perform a mod to the Cheyenne adding a drive adjustment pot so I can peak drive and then turn it down below 3ma.  It just involves replacing the spot switch with a pot/pull switch and a simple transistor circuit.

Having a ball with the old twins!