Rega Research Factory Tour Part 4

In this "episode" you will see that Rega's dedication to quality control extends from the top of the line P10 to the very bottom P1.

All get similar testing. The testing quality itself has improved over the years too. Co-founder Roy Gandy also shows off bearing quality. There's a lot to watch here, and in the next "episode" as well!

COMMENTS
Hergest's picture

I'm thoroughly enjoying all these videos so many thanks for putting them up. The tolerances on the speed and the testing by Rega throws up a lot of questions though such as why does nearly every reviewer mention that Rega engineer their turntables to run fast and when they measure them using an app do indeed find they are fast. Mr Fremer included in this review of the Planar 3.

http://www.analogplanet.com/content/rega-gives-popular-3-series-new-spin...

It's clear from the video that Rega do no such thing so where are the discrepancies coming from? Are reviewers faulty in their measuring? Rega's measuring seems incredibly accurate, certainly far more than a test record and an app.

Michael Fremer's picture
There Planar 3 has an acceptable speed variation range wider than that of the more expensive turntables and narrower than the less expensive ones.... it's shown in the chart.
Hergest's picture

Indeed. It was that I noticed and the subsequent speed test that Rega did that lead me to post as clearly Rega do not engineer their tables to run fast yet reviewers and forum posters still insist this to be the case. In the post I linked to about the Planar 3 the US distributor posted that he tested the speed of the table before sending it to you and he measured it as being 33.3 exactly yet you got a slightly different result. The remembrance of that review and then seeing the Rega measurements make me wonder about the speed testing methods employed by reviewers. Is the test record manufactured to 100% accuracy for its test tones? Is the sound recorder that measures the subsequent test tones calibrated perfectly? It might only be a small difference but it has lead to a slur on Rega and their engineering as regards speed. I am very glad that you have shown the level of testing the company do.

rtrt's picture

I'd say yes the chart does illustrate that their tables are engineered fast.

I assumed that on the chart - green is acceptable/pass and yellow is somehow unacceptable/fail but less so than red.

Based on that - I'd say a passed table is exactly on 33.333 speed or faster up to the listed max for that model i.e. their acceptable tolerances mean that many tables they supply will be fast by design.

So if you take a sample of say 100 tables, then some % will be exactly on speed, the majority will be fast to some extent and a a very small number (maybe none) will be out of tolerance i.e. slow or too fast beyond the tolerance.

Even the table they tested and showed as a pass was clearly slightly above 33.333.

If yellow means acceptable (though what does yellow mean if not somehow less acceptable than green), then its clear that on the cheaper tables, there's quite a bit more tolerance for higher speed than lower speed. Again dictating that there'll be more fast than slow versions of the cheaper tables shipped - everything else being equal.

Anyone else put a different interpretation on it?

Mazzy's picture

I love my RP6 but it does run at 33.7. I'd say that's a bit fast.

cdlp4578's picture

The displayed QC tolerances are purposely chosen to let fast ones through and slow ones not. I suppose it is a chicken/egg question as to whether the tolerances came first or if it was a result of aggregate reject data and choosing tolerances afterwards, but the bottom line is Rega clearly displays what they allow and don't allow through their QC. Not suggesting it is valid, invalid, right, nor wrong. But it is open and clear - and that is valuable to customers and potential customers.

Speaking personally, I am not thrilled about tables running more than .1% off spec, but I also cannot tell you any other company's tolerance levels other than Rega's. That more than anything else gives Rega a leg up on the competition whenever I am looking at a subsequent purchase.

Michael Fremer's picture
I used the Feickert test record, which Rega had on hand and they confirmed its accuracy and that it was a well-engineered test disc. If you look at the Planar 3's allowing speed variances, the tested turntable was within tolerance. The odds that it ran at precisely 33.3 are not high.
theboogeydown's picture

Michael,
I was left curious, did you ever got the opportunity to ask Roy how Brexit might effect their business going forward? They appear to be the quintessential company, caught in it's cross hairs. Not as interested in the political so much as the economics.
Thanks

TerryNYC's picture

Thanks for this excellent series. There really are no substitutes for physics, engineering, and modern manufacturing disciplines. I'm hoping you will cover the Naiad TT & arm even though these are apparently still prototypes only. As they are Rega's cost-no-object solutions based on 40+ years of R&D and product evolution they are bound to be interesting.

Ortofan's picture

... the identity assigned to Patrick McGoohan in the '60s UK TV series 'The Prisoner.'

bent river music's picture

one commenter may have hit the nail when he mentioned that a larger proportion of tables are fast - as a former QA person myself sometimes manufacturers will bias the distribution one way to minimize the number of parts that fall below the intended target to avoid what they perceive as more undesirable quality than being above the intended target - an intentional bias in other words.

atomlow's picture

Hi Michael,

I'm not sure if you go back this far and answer questions but watching the video I noticed the wow and flutter charts. The RP10 has a wow and flutter tolerance of +/-0.1% and turntables like the p3 have a tolerance of +/-0.2%.

If you go and look at the Pro-Ject Debut Carbon specs, for example, http://www.project-audio.com/main.php?tech=debutcarbondc&cat=turntables&... the wow and flutter tolerance is +/-0.1% which is the same as the RP10? Is this possible?

Dr. Feickert says Feickert turntables have wow and flutter tolerances that are +/-0.06%

Do Rega turntables have more wow and flutter than most other turntables manufacturers or are manufacturers not giving correct specs?

I've noticed you updated your PlatterSpeed app to include wow and flutter. I'll be watching new turntable reviews coming in to see how they really test.

Thanks!

X