Project Start - 20141110

(MOSFET) power amplifier #1 - 20141110

       PROJECT GOAL: Create several amplifers capable of driving the deflection on the Cyclops TV monitor. These amplifiers will sit between the Tiny TV Mk2 board and the Cyclops. An X\Y voltage will be converted to a current of several amps. They must pass DC to 300 or 400 KHz.

Amplifier #2: National Semiconductor LH0101CK high speed power amplifier

       Let's begin by looking at idealized amplifier, the LH0101CK power opamp. From the [Texas Instruments datasheet]: The LH0101 is a wideband power operational amplifier featuring FET inputs, internal compensation, virtually no crossover distortion, and rapid settling time. These features make the LH0101 an ideal choice for DC or AC servo amplifiers, deflection yoke drives, programmable power supplies, and disk head positioner amplifiers. The LH0101 is packaged in an 8 pin TO-3 hermetic package, rated at 30 watts with a suitable heat sink. 30 watts per channel? Wow.

       The first photo above shows a second possibilty. A discrete component amplifier based on MOSFET power transistors. An opamp provides input signal conditioning and DC offset control. The MOSFETs provide large current gain, hopefully, with a usable bandwidth. The current opamp, TL072AC, is not fast enough for horizontal scanning ramps. It should work OK for vertical scan. The TL072AC opamp will eventually be replaced with a video speed opamp. This amplifier is a work in progress.

       A third possibility is a flyback scanning type of generator as well. Just for completeness.

Project Update - 20141112

LH0101CK based amplifier PCB vs the McGyver Method - 20141112

       An LH0101CK based amplifier could be built on a manufactured printed circuit board or brought to life on a simple lab breadboard. My CAD designed PCB has banana jacks for power input and output. They must transfer a lot of current with minimal loss. The breadboarded amplifier on the right will be for testing ideas. The printed circuit board may be made next month depending on my funds.

Schematic: LH0101CK amplifier PCB - 20141112

       The above schematic is for a two channel system to follow the Tiny TV 2 board. The first two opamps, TL072AC U1-A and B, is for adding negative DC offset to the incoming unipolar, 0 to +5V, ramp signal. The centering pot drives the negative input of the amplifier with a positive voltage that has the effect of pushing the output voltage in the negative direction. Pots R7 and R16 are the raster centering controls. Scan size is adjusted on the Tiny TV rasterizer board. Centering is adjusted on this board.

       The outputs of the U1 opamps drives the two power opamps, U2 and U3, and finally, the deflection yoke coils. Rsense develops a voltage proportional to the current flowing in the yoke. This provides the feedback to the power opamp, closing the gain loop. Rdamp is a damping resistor. This part is chosen empirically to lower the Q of the deflection yoke just enough to quench any ringing, a form of sweep distortion.

Project Update - 20141113


       Actual testing has begun. Weird stuff was expected and weird stuff is happening. For instance observe a very good vertical waveform. Only when driving the horizontal windings. The yoke does contain some internal resistors and capacitors (snubbing networks). I'll eventually try removing or changing those. We are able to generate a very good ramp current as seen on the scope. This is running +/- 200mA as measured with my Tektronics A622 current probe set to the 100mV/A scale. Note that the wire passes through the probe twice because of the loop. This doubles the measuring sensitivity.

       Testing this evening is with my discrete MOSFET power amplifier running from +/-12V. A schematic of the amplifier will be uploaded tomorrow evening if time permits. It is not very complex. That comes as we try to add the necessary speed (video opamp to replace an MC1458 or TL072) and equalization to get the bandwidth up into the hundreds of KHz. It currently might be a pretty good audio amp for all I know.

Project Update - 20141114

Schematic Diagram: MOSFET Power Amplifier - 20141114

       A basic text book totem pole amplifier. The main function of this configuration is to change the incoming zero to five volt ramp signal into an alternating current through the inductor. Preferably with a high degree of fidelity. The first test of the amplifier uses a generic MC1458 opamp chip. This does not have a lot of bandwidth to begin with. The amplifer target bandwidth is no less than 300KHz. It requires that much to pass the horizontal ramp signal with minimum distortion.

       The first step to increase bandwidth is to replace the opamp with a video speed opamp. The faster the better. There are some 85Mhz video opamps coming in the mail. I may also obtain some dual opamp versions of the quad current feedback opamp I used in a video gamma circuit. These are rated at 200MHz and having a current input in this situation just might rock.

       If necessary, the MOSFETs could be connected to a much higher voltage power supply than 12 volts. I have material on hand to make a +/-30V supply if neccesary. The higher voltage makes more current flow in the inductor - right now! This will only be necessary if I prove in testing that the 12 volt version is actually current starved.

Project Update - 20141116

Two axis raster scanning. Bad raster sanning. But, raster scanning none the less! - 20141116

       Placing one of the LH0101CK amplifiers on the same board with the MOSFET amplifier. Since the MOSFET worked well for vertical scan, I figured I'd try the power IC for horizontal. Initially had picture about a half of an inch wide. 24Vp-p at a half inch worked out to 250Vp-p for approx full screen. This makes sense as the yoke I am using is for a vacuum tube TV.

       Wondering if a step up transformer would even work, I wired one into the circuit. A 50Hz, 220V to 12V iron core transformer. I was amusing my self. Certain of the failure. What? It worked. Scan waveform is just as hosed. But, size is what I'm trying to increase and increase it it did! Now to find a wideband 15:1 step up transformer. The vertical scan is capable of near full screen height. It is reduced here to make the image as 4:3 as possible.

       On a second whim, I plugged the five volt video into the Cyclops blanking input. That also worked great!!! It only needs 5 volts to make full brightness and contrast video.

       The LH0101 runs hot as a pistol without a heat sink like that. But, it is operating OK. Not perfect yet. But, the part proves to be functional. For sanity I tried two of them and got identical results. So neither of those parts sseems defective.

       I will post scope shots of the waveforms later. The vertical is good. The horizontal is generating parabala waveforms. This can be seen as the image fold over in the pictures. The image rotation is the result of my digital camera. Because of a repair, it does not sit level on the tripod. I usually derotate my pictures. But, wanted you to see these as unretouched as possible.

       This is an on going project. Check back every three or four days to follow the progress.


       1. PDF file [LH0101] 60W high power opamp


        Created: November 10, 2014 Last updated: December 1,2016