4QD series controllers.
Issue number history
Contents, this page
Warning: this listing was originally generated for 4QD internal use only, so is not as fully detailed as it should be if it were intended for general circulation. Also 4QD's policy is one of continual improvement, so changes tend to be small and frequent. Also changes usually first appear as a board modification on an the earlier board issue! This makes change history difficult to document fully, but it does improve the product! The history is however included here as it may be useful! No liability is accepted for the accuracy of the information in it!
The 4QD series controllers are in two parts: the top, control board and the base board to which are mounted the MOSFETs and busbars. There are two base boards: one for the 4QD-150 and 4QD-200 series and a separate one for the 4QD-300 series.
Broadly speaking the two boards (base and control) are compatible throughout their release history but some features that have been added affect both boards so great care should be exercised in swapping boards! As a general rule do not exchange boards without written confirmation from 4QD.
In particular, do not mix boards of issues 1-5 with board of issue 6 or later. Boards of issue 5 or earlier are not supported by the factory and service information for these is limited. Replacement boards of latest issue can be supplied and these will fit the old metalwork.
Replacement boards are available. Contact us for prices.
Control Board
Bare board part number: 30-15
Assembled board number: 80-15
Generally the board is very much the same for the -150, -200 and -300 series. However the values in the current limit circuit have been changed from time to time to match different MOSFETs and different value are used in the -300 to make the current limiting more nearly symmetrical in forward and reverse..
There is also a resistor change (two resistors) in the armature voltage sensing circuit for -24, -36 or -48v versions.
The following issue numbers have been released.
- Issue 01 October 1991. 3524 version. The 3524 is not a good chip for motor control, for reasons see PWM motor speed control circuits.
- Issue 02 September 1992. 339 version. Top board redesigned using LM339 modulator.
- Issue 03 March 1993. 339 version.
- Issue 04 July 1994. Mk IV, new circuit. Completely re-laid. Main change is to the acceleration and deceleration ramp circuits.
- Issue 05 October 1994. Mk IV revised, minor mods
- Issue 06 May 1995. Main improvements are:
- HPLO recovery clamp (diode to pin 14 IC4). Without this the high pedal lockout could take a long time to reset, as the controller internal demand speed could be at full speed and had to ramp down to zero before resetting.
- overvoltage clamp. Fitted to limit the maximum regenerated voltage - without this a destructive overvoltage can occur if the batteries fall off or get disconnected as the vehicle is in regen braking mode! Not often a problem!
Requires issue 4 (or later) baseboard.
- Issue 07 November 1996
- 470p on demand speed line. Some controllers had a small high frequency instability problem.
- Issue 08 July 1997 main improvements are:
- pin 9 of 14 way connector now used for undervoltage limit. Pin 9 was unused previously. Feature requires base board issue 9 or later.
- brake defeat link. Bridging this inhibits regen braking mode.
- 10µ added b-e of Tr1, the pot fault detect transistor. Under certain conditions, extreme noise on the pot can cause a problem.
- Issue 09 April 1998 main improvements are:
- RAT modified. Reverse Acceptance Threshold is the (adjustable) motor voltage at which it is considered to be stopped for the purpose of reversing or applying the parking brake. The threshold minimum was found to be too high on some 48v controllers and on some rail vehicles (with low rolling friction).
- c/l bounce clamp fitted (needs modification). The controller is closed loop, with armature voltage feedback. For fast operation the current limit works direct on the modulator, i.e. inside the feedback loop. In certain applications (go-karts) a bounce was noticeable as the current limit disengaged and the feedback loop recovered.
- Issue 10 October 1998 main improvements are:
- c/l bounce clamp fitted fully!
- Footbrake inhibit input modified for hire vehicle use.
- Issue 11 April 1999
- Pin and header links fitted for mode selector (single ended/joystick) and for half speed reverse.
- 1K fitted in series with reverse input: there is a 100n decoupling capacitor on this input and on higher voltages the capacitor charging surge current caused problems.
- Issue 12 July 2000
- 10n capacitor added as a modification, during production, on decel ramp: MOSFETs get better and there is now sometimes a magnetic field interaction between the MOSFET current and the decel circuit.
- Break point added between ramp and modulator: this is for possible insertion of the tacho generator feedback system.
- Issue 13 February 2001
- Diode added on HPLO: one customer found be could defeat the high pedal lockout by switching the ignition off then on again quickly!
- 10n cap added on deceleration preset.
- Transistor identification numbers added on the board silk screen.
Early issues (above) are now too old for factory service.
- Issue 14 May 2002
- The only change is the addition of space for another resistor in the current limit circuit: this is preliminary to changing the base board to make the controller current limit symmetrical (equal forward and reverse currents).
- Issue 15 November 2003
- Two 150K resistors changed to 100K: these are in the joystick area of the circuit and 100K now gives better centring than the original 150K.
- Option added to switch in or out the deadband at zero speed. This can be added as a modification to earlier boards.
- As from April, 2004 controller current limit is now symmetrical. Reverse current is the same as forward current. See 4QD series controllers - Current limits for details.
- Issue 16 December 2004
Two minor changes, to prevent a couple of very obscure faults. For all practical purposes, 16 is the same as 15 - except for the addition of the HPLO defeat.
- HPLO defeat link added. If these pads are bridged out with solder, HPLO is defeated. See Disabling High Pedal Lockout function for earlier issue controllers.
- 47K added Tr 2 base-emitter. Very occasionally a controller gets stuck in reset at power up, and won't start up. Details of the extra resistor.
- Resistor added in reversing latch. We had a couple of controllers which exhibited a problem in the reversing circuit (occasional 'stickiness' at zero speed when going from forward to reverse).
- Issue 17 July 2005
- Half speed reverse now operates in the armature voltage sensing rather than in the demand speed. This makes 'half speed' more accurate across the demand speed range.
- Pot fault level increased: because of component value changes, the level would sometimes not allow a 22k pot to be used.
- Joystick dead band increased: some sticks would not give zero speed within the controller's deadband. This modification was also fitted on some issue 16 boards (it's a resistor value change).
- During the life of the issue 17 (December 2006), we have had to alter the decoupling: as MOSFETs get 'better' their gate capacitance increases and this was causing feedback via gate charge current spikes - this can cause an audible squealing from the motor under certain conditions. It's a particular problem on the 4QD-300-48, but the mod has been done to all production.
- Issue 18 June 2007
- The modification to the decoupling is made permanent.
Base Board - 150/200
Bare board part number: 30-16
Assembled board number: 80-16
Warning: the base board values vary significantly for the various operating voltages.
The following issue numbers have been released.
- Issue 01 October 1991
Issue 1 had a single 3 pin connector for ignition and parking brake.
- Issue 02 November 1992
Ignition connector moved. Now 2 off 3 pin connectors, one for brake, one for ignition
- Issue 03 April 1993
Ignition changed from 3 pin to modern 4 pin to simplify external wiring.
Issues 1, 2 and 3 are very early production: generally they should be replaced with a later issue board. These boards are so old that our internal records are sketchy, having been crippled by an early disc failure!
- Issue 04 July 1994
main improvement is a new parking brake driver. The brake connector was moved, now adjacent to ignition near power terminals. Over-voltage trip added.
- Issue 05 May 1995
main improvement is overvoltage feed to control board (overvoltage clamping). Requires issue 6 (or later) control board.
- Issue 06 September 1995
Gate clamp zeners, Fuse clips.
- Issue 07 April 1996
Hiside c/l and antiphase switching. New thermosensor, 95 degree (was 70 degree).
- Issue 08 October 1996
Brake current limit/trip fitted.
- Issue 09 September 1997
Undervoltage cutback fitted.
- Issue 10 April 1998
Option: parking brake may now be converted to brake light driver (factory option). The normal parking brake driver was also modified to include a delay as the earlier circuit could give nuisance tripping on noise transients.
- Issue 11 15 March 1999
Modification to parking brake driver to latch it, so that it does not disengage if the vehicle is shunted whilst on a hill. This modification is added to some issue 10 controllers. Pullup MOSFET driver converted to complementary Darlington.
- Issue 12 22 June 2000
10n caps added motor outputs to earth: this is to help prevent the occasional motor noise spike causing seemingly random MOSFET failure. The phenomenon is pretty rare however.
Sometime in this issue, the resistor in series with the overvoltage clamp was reduced from 3K3 to 1K0 to flatten the clamping and improve transient clamping.
Adjustable under-voltage protection fitted as a modification. See Undervoltage modification - this can usefully be retro-fitted to -300 controllers, but made it into full production initially on the 150/200 series.
The parking brake driver was also modified: earlier controllers momentarily activated the parking brake when they were powered up.
As from April, 2004 controllers are now symmetrical. Reverse current is the same as forward current.
We regret that we no longer offer service facilities on base boards issue 12 and earlier. Replacement base boards are available.
- Issue 13 July 2003
Undervoltage protection preset fully incorporated.
- Issue 14 September 2004
Minor changes
- Issue 15 November 2005
More minor changes. Issue 15 now has solder resist on the board so is not dip varnished in production.
- Issue 16 July 2006
Board is not with plated through holes: no difference to the user, but simpler assembly
- Issue 17 July 2007
Large decoupling capacitors have been replaced by a lot of smaller ones. In some high current situations these capacitors could overheat: the many smaller ones have better performance in such 'overload' situations. It may also allow us to offer higher voltage (60v, 72v) versions in the future.
Base Board - 300
Bare board part number: 30-28
Assembled board number: 80-28
- 01 First version.
Had 3 pin ignition connector. Subsequently changed to 4 pin (same as 4QD-150 and 200).
- 02 Second issue: board has all the features of standard base board issue 11.
- 03 March 2000
Hole pattern changed for PNP drivers: we had already been using BC339 which had a different configuration. Current source - available current increased.
We regret that we no longer offer service facilities on base boards issue 11 and earlier. Replacement base boards are available.
- 04 April 2001
Generally, only minor changes. Sometime in this issue the resistor in series with the overvoltage clamp was reduced from 3K3 to 1K0 to flatten the clamping and improve transient clamping (we had a few failures on test, during regeneration, from a mains power supply).
During this issue, variable undervoltage protection was added as a modification.
As from April, 2004 controllers are now symmetrical. Reverse current is the same as forward current.
- 05 June 2004
Undervoltage preset added
Several minor modifications.
- 06 June 2006
Minor changes. Main change is extra decoupling on internal 12v line - some of the modern MOSFETs have very high capacitance and this can cause problems without the decoupling changes.
Page Information
© 1998-2008 4QD
Page's Author: Richard Torrens
Document URI: www.4qd.co.uk
/serv/4qdhist.html
Last modified: Monday, 19-May-2008 10:01:34 BST