MRF150 Bias

[KA3EKH]

I have a Harris RF-1110B 1.0 kW amplifier that has four panels each with four MRF150 devices on each panel, when I got it several of the MRF150 devices were shorted. Been using it for running teletype at around 350 watts just running panel one and two, with panel three and four pulled. I did isolate the drive to the two missing panels and have been using it that way for over a year now. Just ordered a bunch of MRF150 devices and want to repopulate panel three and four. Little background, there are two regulated high current fifty volt power supplies with one driving panel one and two and the second one powering panel three and four. On each panel there is a bias control circuit for each device, along with a switch to kill bias for any one device and a pot to set the bias current to that device. I have built up a external box that will allow a panel to be operated outside the amplifier under DC power, no RF and it has metering so you can kill the bias on any or all devices and adjust the bias one at a time to balance them.
The problem is that I have a box full of used MTF150 devices and once I plug them in to the panel was going to do the bias adjustments but wanted to know if there was anything else I should be doing in the process. Figure I can reduce drive and run everything at a couple hundred watts to start out. Each device has a 0.3 Ohm resistor in series with its drain before its output transformer and can balance by reading current that way but will balancing current at DC be ok at RF?

[DMOD]

I have a Harris RF-1110B 1.0 kW amplifier that has four panels each with four MRF150 devices on each panel, when I got it several of the MRF150 devices were shorted...

A schematic of this beast might help us to determine what you're asking.

Phil - AC0OB

[KA3EKH]

It’s a little difficult being the manual spreads it across four pages, one for the bias networks, one for the input splitters, one for the amplifier devices themselves and the last for the combiners and sensor stuff.
Long story short the four devices on each panel are deployed in pairs with each device having its own self contained bias control. That makes eight sets of MRF150 devices, sixteen bias pots on four panels developing 1000 watts; Let me try to attach a block diagram.

[DMOD]

With the new 150's in there adjust the DC bias of each amp according to spec since this bias should place the Mosfets in their linear range of operation.

Starting at a low RF power input, key it and check the temps of the Mosfets; repeat at each increase in power step.

Now this is assuming you have a stable and low ripple power supply. How old are the electrolytic caps in each power supply?

 

[KA3EKH]

Built up a test jig for sorting thru the bunch of MRF150 devices I have here and thought I would put up a schematic in case anyone else is interested. With two power supplies, one for the drain and a variable for the gate you can get an idea for the current gain by comparing gate voltages for a given drain current. Also lets you know if the device is shorted because the light will light brightly or if the device has high leakage from drain to gate or is open. The switch forces the gate to ground to see leakage and cut off the device under test, also I feel better killing the bias line before messing around with the gate line anyway.
It’s built on a old amplifier from a MRI machine, who would have thought that a MRI would use both MRF150 and MRF148 devices? Bought the panels to strip the devices but they also provide a good source for the jest jig.
If there is additional interest may have to do a video!

[Opcom]

Built up a test jig for sorting thru the bunch of MRF150 devices I have here and thought I would put up a schematic in case anyone else is interested.

That looks very practical! Thanks!

[DMOD]

Built up a test jig for sorting thru the bunch of MRF150 devices I have here and thought I would put up a schematic in case anyone else is interested. With two power supplies, one for the drain and a variable for the gate you can get an idea for the current gain by comparing gate voltages for a given drain current. Also lets you know if the device is shorted because the light will light brightly or if the device has high leakage from drain to gate or is open. The switch forces the gate to ground to see leakage and cut off the device under test, also I feel better killing the bias line before messing around with the gate line anyway.
It’s built on a old amplifier from a MRI machine, who would have thought that a MRI would use both MRF150 and MRF148 devices? Bought the panels to strip the devices but they also provide a good source for the jest jig.
If there is additional interest may have to do a video!

Look at the curves on page 4

https://cdn.macom.com/datasheets/MRF150.pdf

and see how they compare to your plotted data points using your test setup.

Phil