Hi Folks. I've designed an H-bridge driver using LM311's. It works well with dcc++ and jmri. I've made KICAD files and produced PC boards and the circuit has been used with straight transistor drivers for a programming track buffer, power darlingtons as a booster, and "external" discrete device darlington outputs and they all have worked nicely. I've also tested the board with Dani's loko-net (sic) arduino code and it likewise works well. I've a short writeup here: https://dcc.wraith.sf.ca.us// Spike Maul
Spike, very interesting. How much of this depends on 5v signals. I'm heading down the 3.3v ESP32 path, so while I can provide 5v VCC, the signal pin max is 3.3v. Also, I was googling the various transistors you note in your write-up, if I wanted the 8A 2n6041/2n6044 where would they go on the board? I'm an ME, not a EE, so am a bit limited in my electronic knowlege
Hello bocabob--the H-bridge "should" be ok with 3.3v PWM and enable signals. I've not tested it but the inputs are simple 2n2222's, not logic gates, so there should be no problem. "should, cough" The output transistors should be selected for their power rating. That "8 amp" rating of the 6041/44's will not be possible in a 15volt circuit, since it exceeds the power rating of the devices. I rate that pair as good for 5 amps/15 volts. Another set might be tip100/tip105. The devices were laid out at the edge of the board so that I could more easily provide adequate heat sinking. Especially, power darlingtons do heat up. INSURE ADEQUATE HEATSINKS. A small fan, such as out of a scrapped AT power supply, or an old computer case fan, should be part of an installation of this board with high power transistors. Installing output transistors: With the input connector close to you, the left side has (reading up) bce/bce and the right side ecb/ecb (always double check). The power input are the right two vias in the center top, and the track output are the left two vias. I've installed a couple of the screw terminals to do testing, but I prefer to "hard solder" if I have a choice. I've been learning these lessons since the 70's so sometimes I take knowledge for granted. Spike Maul
Spike, got it, watts on the power dissipation rating. I take it from your post that the power rating (which I assume is for the pair) is the rating for the 4 transistor bridge and is not the sum of the two pairs? so Tip 100/105 is 80W which is about 5A for a 15V HO track Tip 110/115 is 50W, ~3A @ 15V Tip 120/125 is 65W, ~4A @ 15V Tip 140/145 is 125W, ~8A @ 15V Yes, I think a fan header would be a required addition to the board!
In looking at the BOM to build this, could you provide ratings for the below? Ref Qnty Value C101, C102 2 47 C103, C104 2 .1 C105 1 .22 D101, D102, D103, D104 4 DIODE Device Diode R101, R102, R103, R104, R105, R106 6 620 R107 1 .1 R108, R109 2 300K R110, R113 2 510 R111, R112 2 39K R114, R115 2 15K R116 1 1K R117, R118 2 4K7 R119, R120 2 470 sorry about the formatting but all my efforts to do a table have failed
The c101/102 take a beating in the thor circuit. They should be low ESR aluminum electrolytic of 2000 microfarad, 25 volts, 105C rating. The 85C rating will "work" I suppose. I used the 47uF/35v solid tantalum in the '6041/44 darlington and they got warm. At first I had 22uF tantalum in there, and they got quite hot. Finally I piggybacked some 2200uF/25V low ESR aluminums in there and that made the power circuit purr. BTW I powered it up with *no* capacitors there, and all the decoders laughed out loud. Literally, they all blinked their headlights. The .22 and .1 capacitors should be 50v. Resistors 1/4w 5%. Current sense resistor .025 ohms several watts non-inductive for the power driver. As for the diodes, a set of 1n5403's worked for me. I also used mr851's. Freewheeling diodes, peak forward current surge 100 amps. Use of 1n400x devices is not recommended. Spike Maul
Last night I had a power darlington block under test. This unit was built up with just a 1/4 watt resistor lead as the "current sense" "resistor." When I opened up a "display dcc++ traffic" window in jmri, I noticed that good current sense readings were being sent to jmri from the arduino dcc++ software. Using the two amp brake sled as a load, I was seeing dcc++ deliver a "15/1024" reading to jmri. Looks like the higher current versions of the lm311 H-bridge driver could simply use a jumper for the sense point. BTW the "track current" window in jmri has not worked for me since jmri4.10.something. Spike Maul