Should I use a Golden Sample or Limit Samples?

Should I use a Golden Sample or Limit Samples

During the new product development phases depending on the type of products that you make engineering teams are faced with deciding whether they should use a golden sample or limit samples, and on many occasions, they are confused about exactly what the differences are between these product samples.

Here we will distinguish each and will help you understand which one you should use…

 

In order to understand which sample you need to use and when, let’s look at what these two different types of product samples are:

What is a golden sample?

The phrase golden sample has many different kinds of meanings depending on what industry you’re in, what type of product you have, and if it’s a whole product you’re talking about or just a single component; but typically ‘golden sample’ means the perfect sample.

This sample is, from a specification, performance, design, and manufacturing point of view, exactly what we want all of our products to be like once they’ve been manufactured. It’s the perfect representation of our product that uses identical materials and manufacturing processes as will be used in mass production.

Golden samples are the culmination of a lot of product development work and several rounds of validation and product builds. For instance, we might make 5,000 units and have all kinds of problems, make modifications and fixes and build another 5,000 in a second run. In this run, the products are 90+% fine and meet the desired specification, quality, performance, and testing requirements; therefore can be considered golden samples. 

There are three key points regarding golden samples to focus on:

  1. These samples show that they meet all the specifications requirements from all different departments and therefore everyone is aligned on the samples meeting the requirements. So when the golden sample meets all your requirements (including performance, specification, manufacturing, quality, reliability, and testing) it means to you that all the manufacturer’s stakeholders, partners in manufacturing, design, and development, are aligned and that the NPI process is running smoothly.
  2. It demonstrates that any design, development, component, supplier, or process changes that were supposed to be put in place in order to achieve the ultimate goal of developing a perfect product have been done correctly.
  3. When you have a golden sample it means that from a quality and manufacturing point of view you met your goal of building a product as it was designed. (Note that it’s not necessarily from a reliability standpoint because those samples will need to undergo reliability testing to assess that they will be reliable enough once in the field.)

Further reading: Learn more about golden samples in ‘Golden Sample: Why You Need It Before Mass Production Starts

Can we use only a golden sample to validate products?

If a golden sample is the perfect form of your product, why would any other kinds of samples like limit samples be required for product validation and testing? Surely a golden sample is fine?

The answer to those questions is, yes, in some cases such as when buying standard off-the-shelf components or products, sometimes a golden sample alone is fine. If your product or component has no variation, then a golden sample can be used during inspection to simply pass or fail it with a yes/no response to ‘is it the same as the golden sample.’ Another example where a golden sample is suitable is for RF-related electronic products (radio frequency like wi-fi, Bluetooth, mobile phone signal, etc). If you don’t have a golden sample your RF is not going to be on target and you’re definitely going to have huge failures in the field.

However, not every product will be precisely the same as the golden sample, so if there is a slight variation can we pass it? Let’s examine limit samples for the answer…

 

What are limit samples?

Limit samples, also known as boundary samples, are different representations of your product. It is possible that you may have a limit sample that would meet and maybe even exceed the golden sample specification and performance. ‘Better’ or ‘worse,’ the limit samples are not the same as your golden sample, therefore they are units that are not as you expect.

Typically limit samples are used when you can’t have a single golden sample because it’s really difficult to nail down the perfect specification, color, texture, or alignment edges and gaps. They help operators understand what is and is not acceptable outside of the golden sample’s specifications and, in particular, quality inspectors who need to make a call on whether to pass or fail a unit that appears to be very ‘close’ to the golden sample standard.

There are situations where the quality inspection team have no idea which product unit is a pass and which one is a fail. In this case, limit samples can be used as they show variations on the high or low sides of the sample’s tolerance limit. An example of this would be shades of color where one limit sample shows a lighter shade which is acceptable, whereas another shows an even lighter shade that is not acceptable.

Operators and inspectors can refer to the samples to make informed decisions, rather than relying only on a golden sample which lets them only give a yes/no answer to the question: “Is the unit in front of me the same as the golden sample?” Instead, limit samples enable them to answer the question: “Is the unit in front of me acceptable to the customer even though it has a slight variation?”

The truth is, if you reject a lot of units even with slight variations or imperfections, you’re going to bear a huge cost for scrap products. So the whole purpose of the limit samples is to reduce the cost of inspection and rejects.

 

When to use limit samples?

For the most part, limit samples are used for visual inspections, unlike golden samples which are used for parametric measurements in production for checking the performance and testing critical parameters prior to shipping.

Visual inspections are a manual process done by the human eye to verify the test samples and test the units through the production process in order to pass or fail them.

Your in-house inspector, third party inspectors from an outside company, and even production operators who are self-inspecting their work will find limit samples a useful reference resource, so they will often be made available in the factory for whoever requires them during production and inspections.

 

An example of how limit samples helped us save a customer’s project

This is based on a real example project that we worked on at Agilian.

A customer needed the plastic covers of their product to be a certain color. The supplier produced the covers, they went into production, and somewhere halfway during mass production when they had shipped maybe 50,000 units already, all of a sudden the supplier announced a large cost rise for the covers.

The customer rejected this and came to us to ask if we’d be able to source a new cover supplier. The key issue wasn’t finding a supplier, it was finding one that could produce covers in the correct color.

The solution was to provide a selection of limit samples from the new supplier to the customer for them to choose which color came closest to the old cover from the former manufacturer (which was the benchmark color). They were able to indicate which were acceptable and which were not and could go ahead with the new supplier instead.

 

Validating, securing, and identifying samples

Golden and boundary samples are an important part of your quality standard as they indicate what is and is not acceptable to suppliers and inspectors, therefore they need to be identified, handled, and stored carefully.

Validating samples

Over time limit samples will be subject to wear and tear and environmental factors such as the sun, heat, humidity, etc. This can change them from a pass to a fail, so validating them over time is important. One set of samples usually has a finite lifetime.

For example, the sun may fade the color of a limit sample left near a window, making its color too light to be acceptable, or if someone dropped it causing damage and put it back on the table and you’re using that as your guideline to pass a lot of parts, you may have just failed a whole batch based on the now damaged sample!

Storing samples

So, aside from validating that they’re still correct, it’s critical that limit samples are stored securely, and protected from damage and the environment. Keeping them in a locked cupboard or drawer in sealed bags with soft padding if needed will help protect them from damage and the environment.

Identifying samples

Dated project information should be added to the back of the sample, such as which version of the product it is for. This will help you identify which sample is for which project, what date it was produced, and what version by which supplier.
It’s also wise to operate a signout system, where a log is kept of every time a sample is removed for use, by whom, and when it was returned.

Renewing samples

Some companies choose to refresh samples at certain intervals to assure that they’re exactly right, or they will remove all samples with a certain date stamp if they have been informed by, say, the design team that a new version has been made. This is a dangerous situation as if someone says that they will remove the old samples, but they forget as it’s close to the end of the day or the weekend, this could result in the wrong samples being used for validating new products or components coming off the line. Checks and balances need to be implemented to make sure that old limit samples are scrapped as soon as possible if no longer required, they can’t be allowed to be seen by inspectors and operators.

By the way, the same principle applies to golden samples, too. If they’re damaged, too old, or you’ve made changes to the product, new golden samples will need to be produced every time including the change in material, design, finish, color, function, etc. The old ones should be scrapped immediately.

 

Can using limit samples reduce or increase costs?

How you inspect using limit samples can be costly. If you want perfect samples to be chosen then it means that it will be labor-intensive and there will also be a lot of rejects from those inspections because they’re not meeting the perfect sample inspection criteria. If you demand perfection in each product, such as for very high-value items like hand-made watches or sensitive products like auto parts and medical devices, the inspections will cost more as you need to use a 100% inspection where every unit is checked individually.

On the other hand, if you are lenient and can accept more variations of the samples, using limit samples can reduce your scrap costs and shipping delays.

 

Conclusion

Golden samples need to be used in production when you are doing mass production, mass testing, and actual performance and specification parameters are required to validate the product. They may be used for visual inspections if the pass/fail result is a simple yes or no.

However, the limit samples (or boundary samples) are typically used for visual inspections when it is really difficult, due to the variability of the product. The alternative is to have several different variations of that product (limit samples) to identify and clarify the upper and lower limit that can be accepted during the inspection process, such as: ‘this lighter shade of blue is acceptable, but this shade is too light and would be a failure.’

Get help

Confused? Need help with designing and developing your product and/or manufacturing it? Contact us with your questions and we’ll answer them!

This entry was posted in Product Quality and tagged , , . Bookmark the permalink.

Comments are closed.