Thanks Tony for your suggestions. What I means is: It happen in two distinct times every time when I used the drive to read through a PC with the software IMD to create images of floppy disks 8 ". Not having the correct power connectors i power the drives in this way: + 5vdc and ground coming by a PC power supply and +24vdc and ground from a Chinese power supply. I must have certainly reversed this last one even for a few seconds exchange +24vdc with the ground. Now both spindle motor supplied also with only +24vdc start smoking in the vicinity of the chip LB1620 and cause of cooling you can not touch it as hot it is. If I try to measure the resistance between pins +24vdc and ground of each spindle motor drive i get 1327ohm on FD0 while i measure 19ohm on the spindle motor drive FD1. There are no visible damage nor to the components nor the PCB wires. In both pcbs anyway the chips LB1620 sunburnt from not being able to keep a finger on it. I searched on ebay and the only place that has LB1620 seems to be: http://www.ebay.it/itm/LB1620-INTEGRATED-CIRCUIT-IC-BOX-74-/390955452872?pt= LH_DefaultDomain_0&hash=item5b06c2c9c8 The input's pins of the pcb are : +24vdc, ground, and the rest could be a control. The point is that I have no experience on these components. Replace the chip could make it burn again for my inexperience. Enrico -----Messaggio originale----- Da: cctech [mailto:cctech-bounces at classiccmp.org] Per conto di tony duell Inviato: domenica 1 marzo 2015 17:42 A: General Discussion: On-Topic Posts Oggetto: RE: How to save two Mitsubishi M2896 drive (spindle motor smoked!) OK, what did you do? Seriously, it may help in finding what needs to be replaced. partly a Is there a good reason for not restoring this spindle motor to its original form? The TC9142 chip is a well-known motor speed controller PLL thing, if it's the one I am thinking of it has a pin called 33/45 which indicates where it was intended to be used :-) I've come across it in disk spindle motors, laser printers (the Canon CX uses them for main and scanner motor control, etc). The LB1620 looks to be the motor commutation IC. It takes in signals from 3 (I guess) hall effect sensors, these are just visible, I think, in the track-side PCB photos. The IC then drives the 3-phase motor windings, most of the time they are star connected. You can get data sheets on both these ICs from http://www.datasheetarchive.com The motor PCB seems to have a 3 pin connector on it, I would guess +24V, ground and motor_run. I should therefore be easily possible to run it on its own for testing. I think the first thing you need to do is find that short. The LB1620 is possibly at fault here, maybe desolder it and see if the short goes away. -tony =

On 03/01/2015 11:13 AM, tony duell wrote: My suggestion to Enrico was to pull the driver transistors that switch between the +24 and the motor windings. If his case is usual there are three of these (often Darlingtons). If the board still smokes when power is applied, you know that the fault does not lie in the motor nor in the driver transistors. I received no response, FWIW. --Chuck

On 03/01/2015 11:53 AM, tony duell wrote: I found the datasheet for the LB1620 and apparently the thing is perfectly capable of directly driving the motor windings. I note that a small (10uF to 47uF) electrolytic is specified between each of the windings and +24 Vcc. Were one of those to short, that would indeed have nasty consequences as the output of the LB1620 is via a complementary pair operating totem-pole fashion for each phase. So, in absence of a readily available LB1620, I'd start with the per-winding caps. --Chuck

On 2015-Mar-01, at 9:56 AM, Enrico Lazzerini wrote: I notice in one of the photos you have written 15V for the zener D102, and on both boards they look like they have been replaced. D102 and R112 are a simple regulator to reduce the +24V down for the chip supplies (looking at the photos to trace the circuit). +24 is otherwise used directly only for the motor windings, in part via Vcc1 on the LB1620. The LB1620 datasheet specifies the Vcc2 range for the 1620 to be only 8.5 to 14V, not 15V. Check the Vcc2/zener level to see whether the zeners are good and what voltage they are. I expect Vcc2 (and the zener V) is supposed to be 12V or 9V.

Indeed, pin 4 (Vcc1) should be +24V, this is for the motor windings drive. You need to look at pin 19 (Vcc2) of the LB1620 or the voltage across zener D102. Follow the pcb trace from the junction of R112 and D102, goes to pin 19, looks like pins 9 & 2 as well, and on to supply the other ICs. On 2015-Mar-01, at 1:38 PM, Enrico Lazzerini wrote: