Yes, I get your drift…
One explanation could be purely time-based i.e. the nose gear hesitates always X seconds after the pump starts to run. One could time it in the videos.
However, I have seen various kinds of weird motor failures. I have a bench drill at home, with the usual sort of capacitor start single phase motor, and after about 20 mins the thing would make a funny whining noise and seize up. It was fairly obviously heat related and maybe the expanding armature touched something.
The motor in this gear pump is going to be a common DC brush motor, and I have seen all kinds of weird stuff in those e.g. a short between some turns on one of the coils will make the motor a lot weaker (but draw a lot more current) but only when the whole thing warms up.
The other possibility is a weird failure of the pressure switch. This is a crap quality switch but is used throughout the piston GA retractable scene – presumably because it is “approved”.
I will change the pump but if the issue doesn’t go away I will start looking elsewhere.
The biggest problem is that myself and my engineer are not allowed to work anywhere near where I am based.
I really don’t get this UK BS from hangar owners – a real PITA for sure !
Use an analog voltmeter for measurements on this pressure switch, it will show the spike, a digital meter won’t do a good job for this situation.
However I don’t suspect the switch or a restriction.
On a commutator head the soldering can become soft when the commutator of the small motor becomes hot. It happens as Michael correctly suggested when the motor is not allowed to cool down.
This is quite common on starters as well, where people tend to start for ages, or frequent attempts with a short interval.
The judder only occurs on the nosewheel leg, and it always retracts last (as it did on last year’s video). Therefore, even though the handswing is easy, there is something about the actuator geometry which puts a bigger load on the motor on that leg. Its possible that that load is intermittently sufficient to actuate the pressure switch, causing the motor to briefly cut out and spike when restarting. You could quite easily blank off the nose actuator and see if the same judder occurs with repeated retractions on the MLG only.
As to why it only occurs on the second retraction – as said the motor is hotter, also the fluid has moved around and got a little hotter, also the actuator pistons have moved. Air bubbles going through the pump could also make the motor surge.
Another test would be to disconnect the pressure switch and pull the CB just before the nosewheel fully retracts (but after the judder point). If the judder is the same that would definitely exclude the pressure switch causing the judder.
I’ve found that the length of time between gear motor “blips” to maintain retraction pressure is a good indicator of the condition of the seals and the fluid. However I’d guess you probably change the fluid on a scheduled basis.
Of course as it only occurs on fast repeated retractions you could just ignore it :)
Update:
A new gear pump went in today.
There were two separate factors in play:
The old pump definitely had a shagged motor – it was drawing way too much current at times. I will send it off to a motor rebuilder to have it looked at. Well, I will open it myself first 
The 40A external power unit cannot always supply the instantaneous pump current (even with the new pump) and goes into a brief current limit, which causes the UP relay to drop out, and then re-engage.
Also Socata have been playing silly buggers. They offered a new mounting bracket, €50, but had no idea what it did, so I bought it anyway. The new pump is about 3mm wider (even though it is labelled with the same P/N)
Old:
New:
and sure enough the bracket is required to mount it. But the damned bracket has a load of rivets, the removal of which would take at least a day (partly due to poor access on the other side of them). So we just modified the existing bracket in situ; perfectly neat. Why the pump diameter was changed in such a subtle way, for the same P/N?
Another thing is that you don’t actually get a new pump for the €1300+VAT. What you get is “a pump” but the whole reservoir portion is useless. Well, it would work, but there is no way to access it to check the fluid level You have to transplant the (evidently Socata-fabricated) reservoir from the old pump to the new pump. It takes an hour or two, and you better not bugger up the o-ring because, ahem, it isn’t for sale.
It is obvious what Socata do. They buy in the new pump, remove the reservoir, and send it off to a metal fabrication firm to weld on the dipstick assembly, plug up the central hole, and paint it.
Then they CNC machine up this complicated piece to retain the reservoir on the pump
You need to take a photo of the nipple angles on the old pump so they can be installed on the new pump correctly (no comments please – we did that one today in the hangar 
)
You need to do all this the day before, to give the PR1422 sealant time to go off… tapered NPT threads are less than ideal for an application where the nipple angles need to be adjusted.
This one is strictly for UK TB owning old-timers…. It was fun working in the hangar which many will recognise
What do you think this whole gear pump exchange thing would have cost, all in, if you had just left the airplane at a major maintenance firm (like most people do) and said “fix it”?
What do you think this whole gear pump exchange thing would have cost, all in, if you had just left the airplane at a major maintenance firm (like most people do) and said “fix it”?
About 3000 quid.
And there is little assurance they would actually fix it – because most firms don’t really understand aircraft, especially electrics.
For example one TB21 owner I know was grounded for 6 months because the dealer didn’t know how to adjust the turbo. And more recently another long story from elsewhere trying to get hold of a prop TKS kit…
A few years ago I turned up at Socata (Tarbes – the top people for TBs) and wasn’t able to start afterwards. They “diagnosed” a faulty starter motor, for which I paid about €1400. This was a ~$300 starter which had been sitting on their shelf for maybe 10 years, no lube, and seized up and burnt out within weeks. But the problem turned out to be the battery all along, which did pass the load test (a Gill, no surprise to me). At that point, I fitted the Skytec high speed starter and a Concorde battery, and never looked back. So not only did I get ripped off, but got duff old stock, and the problem wasn’t fixed in the end. This is pretty typical in the trade. Take my elevator trim, freezing up at FL140… The firm which did that is well regarded by their customers, who never go above 2000ft. Sure there are good firms out there but I haven’t come across one yet in my 12 years of ownership.
We did the pump in about 7 man-hours.
BTW I documented it above so others can benefit, not just because I like to have a moan
The knowledge that is was done properly is priceless
I opened up the old gear pump.
There was nothing wrong with the brushes or the commutator. The brushes look like they have 6+ mm on them and the commutator wear depth cannot be felt with the finger so less than about 0.2mm. That should not be a surprise because the total running time on the motor is around 1500 cycles (flights) of about 10 seconds each i.e. 4 hours total running time, which is nothing.
There is a weird bimetallic thermal cut-out, which needs a very high temperature to operate. Even a 360C soldering iron didn’t move it. But it is a prime candidate for causing trouble, looking at the way it is riveted. It could easily be heating up and interrupting the current, briefly. I would bypass this switch, since it doesn’t do anything useful in this application because there is a CB on the pump anyway.
The motor looks like something out of a chinese hoover, although the commutator and the brush assembly are sturdy.
The hydraulic pump portion turns easily with a screwdriver.
So what caused the original problem is a mystery. Except that I did see high current being drawn – 30-40A – during the 1-sec or so pauses on the UP cycle. So it could still be something in the hydraulic pump portion.
Does anybody have any ideas on how this pump could be tested? It needs to be tested under a realistic hydraulic load. It would need to be filled with fluid and run against a pressure relief valve set to something like 1400psi.
This video shows the current drawn
Around 0:25 you can see the new pump pausing on the way UP. It never does this when running on the aircraft battery ( no matter how many times in succession we cycled it) which led me to conclude the pause is caused by the 40A power supply going into current limit and tripping out the UP relay, briefly. What is slightly worrying is that the old pump was doing exactly the same thing, but regardless of where it was powered from.
But there is a difference. During the pauses, the old pump was drawing a LOT of current (30-40A indicated) whereas the new one is drawing nothing. And the only possible way to draw a lot of current is for the pump to be running under a heavy load – or perhaps a faulty motor.
You can tell when the current exceeds 40A because the voltage briefly drops down to about 26V.
Frankly speaking, paying 3000 quid (I guess you mean pounds?) for the repair of a miniature low pressure hydraulic pump seems like total madness. Also, you seem to have lots of time, lots of money and lots of knowledge and the right spirit to fix things. Why don’t you do the experimental community a service and get RV-7 or -8 that you can experiment freely with, possibly making lots of new and nice solutions and products along the way? I mean seriously, there is no need to give up your IFR flying or anything either, people do both.
Anyway, this pump of yours, I guess it is an off the shelf hydraulic gear pump? There are a couple of different main designs of them though. The failure mode is usually they get worn and start to leak internally (they cannot create enough pressure and/or enough flow), but this won’t happen unless it cavitates often or your oil is full of sand or something. It looks to me the load suddenly increases, and this may be a fault somewhere else in the system, some pressure relief valve, adjustable flow restrictor or similar that is worn and erratic. Do you have the hydraulic diagram of the system? and a picture of the internals of the pump?
